View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Institutional Repository for Minnesota State University, Mankato Minnesota State University, Mankato Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato All Theses, Dissertations, and Other Capstone Projects Theses, Dissertations, and Other Capstone Projects 2018 Effects of Integrated and Independent Speaking Tasks on Learners' Interactional Performance Yen Vo Minnesota State University, Mankato Follow this and additional works at: https://cornerstone.lib.mnsu.edu/etds Part of the Language and Literacy Education Commons Recommended Citation Vo, Yen, "Effects of Integrated and Independent Speaking Tasks on Learners' Interactional Performance" (2018). All Theses, Dissertations, and Other Capstone Projects. 836. https://cornerstone.lib.mnsu.edu/etds/836 This Thesis is brought to you for free and open access by the Theses, Dissertations, and Other Capstone Projects at Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato. It has been accepted for inclusion in All Theses, Dissertations, and Other Capstone Projects by an authorized administrator of Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato. Effects of Integrated and Independent Speaking Tasks on Learners’ Interactional Performance By Yen Vo A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Teaching English as a Second Language Minnesota State University, Mankato Mankato, Minnesota November 2018 November 16th, 2018 Effects of Integrated and Independent Speaking Tasks on Learners’ Interactional Performance Yen Vo This thesis paper has been examined and approved by the following members of the student’s committee. ____________________________________ Dr. Glen Poupore, Advisor ____________________________________ Dr. Karen Lybeck, Committee Member i Abstract Interaction has been considered an important element in second language acquisition (Long, 1983). Also, the ability to effectively and appropriately interact with others is one of the important sub-constructs of oral communication (Kramsch, 1986; Ockey & Li, 2015). Researchers in the field of language teaching have raised the importance of how different task types or task formats affect learners’ interaction. As a result, many efforts have been made for a better understanding of what task types/formats are more appropriate for promoting language acquisition as well as interactional ability. However, little has been done to investigate how integrated speaking tasks and independent speaking tasks affect the way learners interact with others. Therefore, this study attempted to examine the effects of these two speaking tasks on how learners interact in pair discussions. A total of 8 language-learner pairs across different proficiency levels who were taking English as a second-language courses at a U.S. Midwestern university participated in the study. Each language-learner pair participated in the two speaking tasks and their performances were transcribed and coded for interaction features based on interactional analysis (R. Ellis & Barkhuizen, 2005). Following this approach, this study focused on six types of interaction features: negotiation of meaning, negotiation of form, negotiation of task content, negotiation of task procedure, negotiation of personal experience, and self-initiated repair (Poupore, 2004; Van den Branden, 1997). The research results indicated that the two speaking tasks were not significantly different in terms of promoting interactional features that facilitate second language learning. However, the independent task, as opposed to the integrated ii task, seemed to produce significantly more negotiation of meaning. The research findings also showed that the learners’ proficiency level might interact with interactional feedback and interactional modifications, with the advanced learners producing significantly more negotiation of content and self-initiated repair compared to their low-intermediate level counterparts. By showing evidence about the effects of the integrated and independent speaking tasks on these learners’ interaction performance, the study helps inform teachers of how different task types may enhance learners’ interactive skills as well as push forward their interlanguage development. iii Table of Contents Abstract ................................................................................................................................ i Table of Contents ............................................................................................................... iii List of Tables ..................................................................................................................... vi Acknowledgements ........................................................................................................... vii Chapter I: Introduction .................................................................................................... 1 Background of the Problem ............................................................................................ 1 Purpose of the Research .................................................................................................. 8 Significance of the Research........................................................................................... 9 Structure of the Thesis .................................................................................................... 9 Chapter II: Literature Review ....................................................................................... 11 Interaction and L2 Acquisition ..................................................................................... 11 Influential Factors on Interaction-driven Learning ....................................................... 13 Task characteristics and interaction-driven L2 learning. .......................................... 13 Learner proficiency and interaction-driven L2 learning. .......................................... 21 Learners’ perceptions and interaction-driven L2 learning. ....................................... 28 Research Questions ....................................................................................................... 32 Chapter III: Methodology .............................................................................................. 33 Research Setting ........................................................................................................... 33 Participants.................................................................................................................... 34 Materials ....................................................................................................................... 35 Tasks. ........................................................................................................................ 35 iv Questionnaires. .......................................................................................................... 36 Data Collection Procedure ............................................................................................ 37 Data Analysis ................................................................................................................ 38 Chapter IV: Results ........................................................................................................ 42 How might the Integrated Speaking Task and the Independent Speaking Task Differ in Relation to the Participants’ Interactional Performance? ............................................. 42 How might the Participants’ Interactional Performance Differ across Different Proficiency Levels?....................................................................................................... 44 What are the Participants’ Perceptions of the Integrated and Independent Speaking Tasks? ........................................................................................................................... 52 Chapter V: Discussion .................................................................................................... 60 Interaction-driven Learning Opportunities across Interactive Tasks ............................ 60 Mediating Effects of Learner Proficiency on Task-based interactions ......................... 66 Relationships between Learners’ Perceptions of Tasks and Interaction-driven Learning ...................................................................................................................................... 70 Chapter VI: Conclusion ................................................................................................. 74 Summary of Research Results ...................................................................................... 74 Pedagogical Implications .............................................................................................. 77 Limitations of the Research and Recommendations for Future Research .................... 80 References ........................................................................................................................ 82 Appendices ..................................................................................................................... 102 Appendix A: Integrated Speaking Task Description .................................................. 102 v Appendix B: Independent Speaking Task Description ............................................... 103 Appendix C: Questionnaire 1 ..................................................................................... 104 Appendix D: Questionnaire 2 ..................................................................................... 107 Appendix E: Transcription conventions ..................................................................... 109 vi List of Tables Table 1 Interaction Variables across Tasks ......................................................................................................................................... 42 Table 2 Interaction Variables across Proficiency Levels ......................................................................................................................................... 45 Table 3 Learner Perceptions across Tasks ......................................................................................................................................... 54 vii Acknowledgements Throughout the process of completing this thesis, I was so lucky to receive a great deal of support and encouragement from many professors, colleagues, friends, and students. Without them, I could not have had the thesis completed. First and foremost, I would like to thank God for his mercy that has always protected and guided me in my life. I am thankful for sending me to such a good university and an excellent TESL program in which I was able to meet so many great professors, colleagues, and friends whose unconditional love and generous support gave me the strength to complete the challenging academic path during these two years. My thanks go to God for sending me good friends and roommates who helped enrich my understanding of other cultures and create long-lasting friendships that will go with me for the rest of my life. Second, I would like to express my deepest gratitude to my research advisor, Dr. Glen Poupore, for his dedicated guidance, insightful feedback, and constant encouragement throughout the study. I also would like to acknowledge his great patience and kindness for spending countless hours guiding me and giving me constructive feedback in every stage of the research even during summer vacation or during his period of recovery from surgery, all of which have shaped this research paper. From the initial stage of forming a research proposal to the long process of IRB application, the stage of presenting my research in GSEA conference till the final stage of editing and publishing my research, all have witnessed his dedication and enthusiasm which inspired me to keep continuing my research. viii Third, I am deeply grateful to my second reader, Dr. Karen Lybeck, for her critical feedback and insightful suggestions which helped improve my thesis considerably. I also would like to express my appreciation to Dr. Sarah Henderson Lee for her generous support to give me access to the potential research site. My gratitude is also directed to the English Department Head, Dr. Matthew Sewell, for generously giving me access to the software Nvivo to facilitate my data analysis. I would like to also thank the manager of Educational Resource Center, Kevin Hebert, for his support in lending me equipment for my research. Fourth, I wish to sincerely thank Alisa Wuest and Nasiba Norova, the instructors of IEP 022 and EAP 125 classes, for kindly helping me reach out to the participants of this research study. I also would like to thank all the students who willingly participated in this study. My thanks also go to Ching-I Hwang and Zhera Shabih for voluntarily piloting the study and giving me suggestion about task implementation. My sincere thanks also go to Sr. Helen Jaeb and Dr. Glen Poupore for their kindness in becoming the speakers in the two videos in this study. Without their support, it would be impossible for me to design the research materials. Fifth, I would like to thank the National Federation of Modern Language Teachers Associations (NFMLTA) and the English Department of Minnesota State University, Mankato, for their generous support for my presentation of the findings of this study at the Language Assessment Research Conference (LARC) at Iowa State University in March, 2018. Having an opportunity to present my research study and learn from other experts at LARC was a great importance for me. Through attending and ix presenting at this conference, I was able to connect with other teachers, academics and professionals who have similar interests as me, but different ideas and approaches. This provided me with extensive opportunities to explore new perspectives in the field. These new perspectives would help not only my future career development but also my future research projects that aim at contributing relevant new knowledge to the field. Finally, I owe a huge debt of gratitude to my family who has been always on my side, given me the best conditions, and encouraged me continuously so that I could overcome challenges to complete my study. My most sincere thanks go to my mom, Hien Nguyen, and my dad, Anh Vo, for always remembering me in their prayers and being a constant source of encouragement and inspiration which helps me follow my passion for studying. I also extend my gratitude to my sister, Quyen Vo, for periodically checking the progress of my study. Without her check-in, I would not have had such motivation to complete the study on time. I would also like to sincerely thank my sister, Sonca Vo, who helped me with issues related to interrater-reliability in this study. I am thankful for my brother, Duy Vo, for giving me useful advice on my spiritual life and encouraging me to overcome the hardship of living far away from family. My gratitude also goes to my aunts and uncles for constantly remembering me in their daily prayers and for sending me warm clothes and nutritious food that keep me survive during such harsh winters in Minnesota. My deep gratitude to you all! 1 Chapter I Introduction Background of the Problem The history of interaction research can be traced back to the late 1970s when Hatch (1978) proposed that, “language learning evolves out of learning how to carry on conversations, out of learning how to communicate” (p. 63). Being inspired by this initial idea, research studies began to seek empirical evidence as to how conversational adjustments led to language learning. Early in the 1980s, Long (1983) introduced the Interaction Hypothesis, arguing that learning happens through comprehension and an efficient form of comprehensible input is the one that has been modified between interlocutors for the sake of message understanding. Later, he further added that not only does modified input increase the comprehensibility of the message, but also provides corrective feedback (e.g., comprehension check, clarification request, confirmation check) from which learners can make salient and available for learning some of the problematic aspects of their interlanguage (Long, 1996). Building on Long’s argument, Swain (1995) stipulated that, besides comprehensible input, it is also important for learners to produce output either in a written or an oral form. This is because in producing output, learners are exposed to opportunities to test their language hypotheses, receive corrective feedback on their erroneous production, increase their metalinguistic awareness, and notice a gap between their interlanguage and the target language when they realize that they are not able to say what they want to say accurately and appropriately. Although noticing does not directly 2 lead to acquisition, it provides a necessary condition – a starting point for language learning (Schmidt, 1990, 2001). Moreover, through the two-way process of interaction, learners can also engage in scaffolding during which the seemingly more proficient learner could help the less proficient person reach a new level of understanding – a concept known as the Zone of Proximal Development introduced by Vygotsky (1978). Therefore, from the cognitive and sociocultural perspectives of language learning, interaction is believed to facilitate the process of second language (L2) learning (Loschky, 1994; Mackey,1999; Pica, 1992; Spada & Lightbown, 2009). Having established the facilitative role of interaction in L2 learning, researchers have since moved to seek empirical evidence that interaction is beneficial to L2 acquisition. One of the earliest studies to examine the relationship between interaction and L2 development is R. Ellis, Tanaka, & Yamazaki (1994), who found that learners who received interactionally modified input were able to show better L2 comprehension and retention of L2 vocabulary than those who received premodified or unmodified input. Similarly, Gass and Varonis (1994) found that interaction had a positive impact on both the comprehension and accuracy of the L2 production of L2 learners who received interactionally modified input from native-speaker interlocutors. The learners were able to perform better in the subsequent trial of the same direction-giving task, with the second trial being five minutes apart from the first one and the only difference between the two trials being the depiction of an outdoor scene. indicating some immediate benefits of interaction to learner production. Mackey (1995) also conducted a research study with five adult ESL learners who participated in communicative tasks with a native speaker. 3 Analyses of the learners’ language production of question forms from pre- to post-tests indicated that learners who were actively involved in interaction were found to show L2 development in terms of question formation. By and large, results such as these and other empirical research studies (Keck, Iberri-Shea, Tracy-Ventura, & Wa-Mbaleka, 2006; Mackey & Goo, 2007; Russell & Spada, 2006) substantiate the direct link between interaction and acquisition in terms of both receptive language skills and language production. More recently, interaction research has shifted its focus to examining the multidimensional construct of interaction and how the manipulation of these constructs affects L2 learning and acquisition. While a large body of research (Abadikhah, & Mosleh, 2011; Adams, Nuevo, & Egi, 2011; Fuji, & Mackey, 2009; Iwashita, 1999; Gass, & Mackey, 2007; Long, 2006; Mackey, 2013; Mackey, Adams, Stafford, & Winke, 2010; McDonough & Mackey, 2008; Nuevo, Adams, & Ross-Feldman, 2011; Sheen, 2008; Swain, 2005) has recently explored how interactional components such as input, feedback, and output opportunities impact L2 development, other research studies (Adams, Fujii, & Mackey, 2005; Lantolf, 2012; Ortega, 2009; Ortega & Iberri-Shea, 2005; Philp, Walter, & Basturkmen, 2010; Tarone, 2009; Ziegler, Ammons, Lake, Seals, Hamrick, & Rebuschat, 2012) have investigated the importance of social, cultural, and linguistic factors on the effectiveness of interaction. With a shift to a more comprehensive research focus, more sophisticated research instruments and measurement techniques would be expected to elicit and analyze relevant data. One of the most common and effective data elicitation tools that meets the demand of current interaction 4 research studies is interactive tasks (Mackey, 2013). These tasks include, but are not limited to, picture description tasks, picture sequencing tasks, spot-the-difference tasks, dictogloss, story completion tasks, jigsaw tasks, problem-solving tasks, and consensus tasks. These tasks can be further classified either as optional information exchange or required information exchange; one-way or two-way; open (also known as divergent) or closed (convergent). Since tasks can be flexibly designed to manipulate the type of input and feedback that learners receive and the type of output they produce, they enable researchers to specify which and how components of interaction influence L2 development. Given the importance of interactive tasks in interaction research, researchers have investigated how different task types and task characteristics affect interaction-driven L2 learning, as well as develop learners’ interactional ability. To date, research studies have found that each type of task facilitates different kinds of interaction-driven learning opportunities. For example, required information exchange tasks tend to be more successful in yielding negotiation of meaning than optional information exchange tasks (Doughty & Pica, 1986; Long, 1981, 1983; Samuda & Bygate, 2008). It has been claimed that closed tasks (only one predetermined answer), with their inherently tight structure, are better for promoting interactional feedback than open tasks (no predetermined answer) while open tasks provide more flexibility and allow learners to pool their language-knowledge resources (Julkunen, 2001; Pica, Young, & Doughty, 1987). A comparison between one-way (one learner holds the information) and two-way (each learner holds different information) tasks indicated that the latter type of task engenders 5 more negotiation and interactional modifications as both learners in a dyad have to take equal responsibility for exchanging necessary information to complete the task (R. Ellis, 2001; Swain & Lapkin, 2000). Furthermore, it has been found that task characteristics such as task complexity, task familiarity, planning time, and task repetition impact the occurrence of interactional features (Bygate, 2009; Foster & Skehan, 2009; Plough & Gass, 1993; Yuan & R. Ellis, 2003). For instance, these research studies have found that repeating the task procedure is conducive to L2 learning since it encourages learners to engage in language-related episodes (LREs), which has been defined by Swain and Lapkin (1998) as, “any part of a dialogue where the students talk about the language they are producing, question their language use, or correct themselves or others” (p. 326). Such LREs are important for language learning because they draw learners’ attention to linguistic forms, whether it is grammatical, lexical, or phonological aspects. It is also suggested that learner-internal factors such as age, gender, personality, L1 background, L2 proficiency level, language anxiety, motivation, exposure to interlocutor’s accent, interlocutor type, interlocutor relationship, individual perceptions, working memory, and so forth, play a crucial role in determining the nature of interaction, interactional patterns, and the type and quantity of interactional feedback (DeKeyser, 2011; Mackey, 2002; Mackey, Adams, Stafford, & Winke, 2010; Mackey, Oliver, & Leeman, 2003; Philp, 2003; Sheen, 2008). For example, it was found that L2 learners may produce more negotiation of meaning as well as other interactional features (e.g., prompts, recasts, repairs) in dyads that consist of two non-native speakers, than 6 when they were paired with a native speaker (Gass & Varonis, 1994; Pica, LincolnPorter, Paninos, & Linnel, 1996; Porter, 1986; Varonis & Gass, 1985). Mackey, Oliver, and Leeman (2003) found that native English speakers (NSs) in both age groups provided significantly more feedback than their non-native counterparts. However, feedback provided by nonnative English speakers (NNSs) in child dyads led to considerably more opportunities for modified output than those offered by NSs in child groups. It is important to know that the study was conducted in a laboratory context, which gives a caution to the interpretation of the result since learners may behave differently in classroom-based contexts. The study also calls for more research in authentic classroom environments to have a better understanding as to how learner variables interact with interaction-learning processes. Another area that currently receives a great deal of attention from researchers is the impact of cultural and social factors on interaction-driven learning. As reported in Philp, Walter, and Basturkmen (2010), participants’ provision of feedback depends largely on the relationships with the partners that they work with. Specifically, it is found that the number of LREs was not quite high when Mandarin learners worked with English students during pair or small group discussions. Retrospective interviews with the participants indicated that the learners rarely provided interactional feedback because they did not want to appear superior or put other learners in embarrassing situations, which often holds true for several Asian cultures that emphasize the notion of saving face. It is also documented that culture plays an important role in shaping the nature of interaction (Adams, Fujii, & Mackey, 2005; Ortega, 2009; Ortega & Iberri-Shea, 2005; 7 Ziegler, Ammons, Lake, Seals, Hamrick, & Rebuschat, 2012). In countries where teacher-centered classrooms are prevalent, learners do not highly appreciate task-based interactions as they have a strong belief that teachers are the only reliable source of knowledge. This negative perception about interactive tasks negatively impacts learners’ willingness to communicate and to provide feedback to their fellow learners during interaction (Mackey, 2013). Despite the increasing recognition of the interplay between the social, cultural and contextual factors on interaction-driven learning, more research on social and cultural factors is needed to provide an inclusive picture of how these factors interact with learner-cognitive processes in the process of L2 learning. One challenge facing researchers who seek to clarify the relationship between interaction and L2 development is that learning is a cognitive process that happens within the learners’ mind. In other words, interactional episodes documented in the learners’ language production may show that learners seem to understand what was not understood. However, examining transcripts alone does not indicate that learning has taken place. Furthermore, research has pointed out that learners “sometimes feign comprehension after negotiation rather than continue to demonstrate their incomprehension to their interlocutors” (R. Ellis et al., 1994, p. 454). It is, therefore, worth investigating learners’ perspectives in combination with their language production to have a deeper understanding of the underlying factors that influence their interactional behavior. With respect to this research focus, introspective measures such as stimulated recalls, interviews, think-aloud protocols or questionnaires are needed because they allow researchers to understand how learners perceive interactional feedback and how these 8 perceptions influence the provision of interactional modifications as well as the subsequent language production. To summarize, previous research has shown that several factors, including task types, learner characteristics, and interlocutor familiarity in terms of the length of their relationship, have an impact on how interactional features are produced in an interaction. These factors seem to be still important in research on task-based interaction to further understand how interaction impacts L2 development (Mackey, 2013). Purpose of the Research In response to the need for more research on task-based interaction, this study was designed to examine how two different interactive tasks (i.e., an independent speaking task and an integrated speaking task) may enhance L2 learning opportunities, and how interaction-driven learning opportunities differ across ESL learners with two different proficiency levels (i.e., low-intermediate and advanced levels). As pointed out by previous researching that negotiation of meaning is not the only interactional feature that facilitates L2 acquisition, this research study considers other types of interactional features such as negotiation of form, negotiation of content, negotiation of procedure, negotiation of personal experience, and self-initiated repairs (Poupore, 2004; Van den Branden, 1997). These interactional features are respectively examined for their frequency as well as for actual modified output. In addition, in considering the influence of learners’ perceptions of tasks on the provision of interactional feedback, the study also explores how learners perceive the use of the independent and integrated speaking tasks in a L2 classroom. 9 Significance of the Research The study is significant to the field of L2 teaching and learning as well as SLA in several ways. First, the study provides empirical evidence about the effects of task types (i.e., independent and integrated speaking tasks) on learner interaction. With evidence about how the two tasks interact with interaction-driven L2 learning opportunities, the study would inform teachers of how different tasks may enhance learners’ interactive skills and push their interlanguage development forward. Second, this study is hoped to provide a better understanding of what constitutes quality interaction and what factors impact the occurrences of quality interaction. Third, the research results can demonstrate how learners’ proficiency levels mediate the effectiveness of these task types on L2 learning. Furthermore, insights from students’ perceptions of their interaction and the use of independent speaking tasks and integrated speaking tasks in L2 classrooms can help language teachers to have a better understanding of how to design tasks that promote learning opportunities while also facilitating negotiation and L2 acquisition. Lastly, the implications drawn from the study can serve as a reference for future research studies which investigate the effectiveness of task types on learner interaction. Structure of the Thesis This thesis is organized into five chapters. The current chapter introduces the background of the problem to be addressed in the study, the purpose of the study, its significance, and the organization of the study. The second chapter reviews previous studies relevant to this study. It begins with recent research about the relationship between interaction and L2 acquisition. Then, factors that influence interaction-driven 10 learning are reviewed, with a focus on contextual factors and individual differences, along with the theoretical background of task-based interaction and their effects on L2 development. The third chapter details the research design, methodology, and data analysis. In the methodology section, participants, materials, and instruments of the study are presented, followed by a detailed description of the data collection and data analysis procedures. The fourth chapter details findings obtained from the data with respect to each of the research questions. The fifth chapter represents interpretations of the research results in connection with relevant literature. The last chapter summarizes the main findings of the study and gives several pedagogical implications of the research findings. The thesis concludes with limitations and recommendations for future research. 11 Chapter II Literature Review This chapter reviews previous research studies that have explored the impact of interaction on L2 learning with a focus on learner-learner interaction during interactive tasks. First, the benefits of interaction towards L2 acquisition that have been documented in a recent body of research are presented (e.g., Brown & Lee, 2015, Bygate, Norris, & Van den Branden, 2012; R. Ellis, 2009; Littlewood, 2007; Van den Branden, 2016; Willis, 1996). Second, the chapter continues with a review of factors that influence interaction-driven learning, detailing how task characteristics and learner-internal factors mediate L2 learning opportunities that happen during task-based interaction (e.g., Williams, 2001; Leeser, 2004; Watanabe & Swain, 2007; Kim 2009). Finally, the chapter ends with an overview of recent research studies on learners’ perceptions and sociocultural factors that impact task-based interaction (e.g., Dörnyei & Kormos, 2000; Schulz, 2001; Mackey, 2002). Interaction and L2 Acquisition As stated in Chapter 1, interactive learning has received a lot of attention from language teachers and researchers for their potential benefits to L2 acquisition (Brown & Lee, 2015). Of all L2 teaching and learning approaches that promote interactive learning, task-based language teaching has become prevalent in recent years as it is argued to provide favorable conditions for interaction, which is likely to facilitate L2 learning and acquisition (Bygate, Norris, & Van den Branden, 2012; R. Ellis, 2009; Littlewood, 2007; Van den Branden, 2016; Willis, 1996). During meaning-oriented interactive tasks, L2 12 learners are exposed to opportunities to negotiate for meaning through different types of interactional features such as recasts, confirmation checks, comprehension checks, and clarification requests (R. Ellis, 2003; Long, 1996, 2006; McDonough & Mackey, 2013; Swain & Lapkin, 2001; Van den Branden, 2006). Through these interactional processes, learners might be able to understand what was not understood, and likely to modify their speech for a better message conveyance. With the potential benefits of interaction for L2 development, together with the increasing promotion of learner-centered teaching approaches in L2 classroom contexts, it is not surprising that learner-learner interaction has received more attention in recent research studies than teacher-student interaction (Adams, 2007; Bowles & Adams, 2015; Chen, 2016; Gilabert, 2007; Kim, 2009; Kim & McDonough, 2011; Leeser, 2004; Riccardi, 2014; Tawfik et al., 2018). Although empirical evidence supports the claim that learner-learner interaction is beneficial to L2 learning, one empirical question of a great interest to SLA researchers is how learner-learner interaction impacts L2 development. In an attempt to address this concern, a substantial number of research studies have examined L2 learners’ oral production during their interactive task performance, and have found that the extent to which L2 learners engage in collaborative interaction and make use of learning opportunities arising during task-based interaction varies significantly depending on several factors. One of the influential factors affecting interaction-driven learning is taskrelated variables (e.g., task design, task implementation, task repetition, task familiarity, task complexity) and another factor is learner-related variables (e.g., proficiency levels, pair dynamics, age, gender, personalities, L1 background, exposure to interlocutor’s 13 accent, interlocutor type) (DeKeyser, 2011; Kim & McDonough, 2008, 2011; Leeser, 2004; Nuevo, 2006; Watanabe & Swain, 2007). Due to the complex meditating effects of these contextual and learner-internal factors on the effectiveness of interaction on L2 development, it is necessary to examine empirical evidence to have a more comprehensive understanding of the relationship between interaction and the acquisition of a L2. Moreover, implications drawn from these studies are of great value to be taken into consideration in the design of future interaction research and instructional programs to facilitate L2 learning and acquisition. This leads to the following section which highlights recent research studies on factors that mediate interaction-driven language learning. Influential Factors on Interaction-driven Learning Task characteristics and interaction-driven L2 learning. Interactive tasks remain the most common elicitation instrument in interaction research as the flexibility of task design allows researchers to manipulate the input and feedback learners receive, the output learners produce, and the degree to which learners make use of interactional feedback (Mackey, 2013). This opens the possibilities for researchers to investigate the relationship between specific interactional components and language learning. As a result, much attention has been given to investigate what aspects of tasks enhance L2 learning opportunities. Research findings indicated that task-related factors, such as task types, task implementation, task complexity, and task familiarity, influence the provision of interaction-driven learning opportunities. How these task-related factors interplay with 14 the process of L2 learning during interaction will be reviewed respectively in the following paragraphs. Before investigating the effects of task types on L2 learning, it is important to know how tasks are classified in the literature. To date, work on task typology suggested that tasks can be classified in different dimensions: (1) one-way and two-way tasks, (2) closed and open tasks, (3) convergent and divergent tasks. With respect to the one-way and two-way distinction, tasks are differentiated based on the reciprocal nature of information exchange between learners (Doughty & Pica, 1986; R. Ellis, 2001). In oneway tasks (e.g., picture drawing), one learner holds the most important information and takes the responsibility for conveying information to the other learner to complete the task. Two-way tasks (e.g., spot-the-difference, jigsaw), in contrast, require learners to interact with each other as each of them holds different pieces of information which needs to be exchanged for a successful task completion. Since two-way tasks involve both learners to exchange information, it is expected that two-way tasks engender more negotiation of meaning and interactional features than one-way ones. Most of the studies lend support to this claim (Gass, Mackey, & Ross-Feldman, 2005; Long, 1981, 1983; Pica, 1987; Pica & Doughty, 1985; Wajnryb, 1990). However, it is not to say that oneway tasks are not valuable for L2 learning as there is evidence that one-way tasks lead to more opportunities for modified output than two-way tasks (Iwashita, 1999; Shehadeh, 1999). Regarding closed/open tasks, closed tasks are characterized by the fact that there is only one correct predetermined answer while open tasks are more loosely structured 15 and have no predetermined answer. An example of closed tasks is a spot-the-difference task while story completion tasks would exemplify an open task. It is hypothesized that closed tasks are more conducive to L2 learning since learners are required to reach a final answer, which leads to more negotiation (R. Ellis, 2003; Nuevo, Adams, & RossFeldman, 2011; Pica, Young, & Doughty, 1987). Julkunen (2001) also suggested that closed tasks are more likely to benefit anxious learners who would feel more secure with a tight task structure. This raises a concern that task types may interact with learnerrelated variables in mediating interaction-driven learning. On the other hand, researchers also lend support to open tasks, arguing that this type of task provides learners with more flexibility and opportunities to manage their own discourse (Leaver & Willis, 2004; Philp & Mackey, 2010). In line with that, Poupore (2004) pointed out that problem-solving prediction tasks, which are more open in nature, seemed to produce more quality interaction than jigsaw tasks, which are more restrictive in their structure and lower the chance for learners to experiment with the language. However, he argued that jigsaw tasks still proved to be valuable to L2 learning as more self-initiated repairs were found in this type of task. Jigsaw tasks also provided favorable conditions for cooperative learning and remained interesting and motivating for L2 learners. Another task distinction, which is closely similar to the closed/open task distinction, is that of convergent and divergent tasks (Duff, 1986). In convergent tasks (e.g., a problem-solving task), learners are required to reach an agreement on a solution to a posed problem while in divergent tasks (e.g., a debate), there is no need for learners to come into a final consensus. Previous research findings have shown that convergent tasks 16 tend to be more successful in promoting interactional modifications and more turns than divergent tasks, although language produced in the latter tasks seems to have more syntactic complexity (Duff, 1986; Skehan & Foster, 2001). Another task-related feature that also attracts a great deal of attention from interaction researchers is task complexity. According to P. Robinson’s Cognition Hypothesis (2001, 2003, 2005), increasing task complexity would promote more interaction-driven learning opportunities (e.g., LREs, recast, self-corrections, and negotiation of meaning). P. Robinson and Gilabert (2007) supported this claim by providing empirical evidence that in comparison with the simple version of a task, the complex version which requires more cognitive demands (e.g., reasoning) leads to significantly more interaction and uptake of feedback, increasing learning from input and the incorporation of form which is made salient in the input. Along the same line, Nuevo (2006) conducted a study on 113 adult L2 learners regarding the influence of task complexity on L2 learning opportunities during interactive tasks. In this study, learning opportunities were defined as the occurrence of interactional features, such as recasts, confirmation checks, and clarification requests. Participants were asked to perform two different tasks (e.g., a narration task and a decision-making task), with each task being designed with two different complex levels. In contrast with the prediction of the Cognition Hypothesis, the findings suggested that different task complexity resulted in different interactional features, but that simple tasks seemed to create a better condition for the uptake of comprehension checks, recasts, and repetition. 17 Building on previous interaction research, Gilabert, Baron, and Llanes (2009) further expanded interaction studies by investigating the role of task complexity and task types in creating L2 learning opportunities. A same group of learners performed three different tasks (i.e., a narrative reconstruction task, an instruction-giving map task, and a decision-making task), with each task being at two complex levels, and data were analyzed based on clarification requests, confirmation checks, comprehension checks, recasts, LREs, and repairs. The research findings indicated that there was a relationship between task complexity, task types, and interactional features. For instance, learners tended to produce more negotiation of meaning during complex task versions. With respect to the relationship between task types and the provision of learning during interaction, the narrative reconstruction task was found to encourage more clarification requests, LREs, and repairs while the instruction-giving map task led to more confirmation and comprehension checks. There was also a higher frequency for repairs in the decision-making task. In the same vein, Révész (2011) investigated the extent to which task complexity influenced learner-learner interaction. Forty-three adult ESL learners from six intact classrooms carried out an argumentative task with three different complexity levels. Data were coded for interactional features and other measures of language production, such as accuracy, syntactic complexity, and lexical diversity. Self-report questionnaires were also used to investigate the modulating impact of learner individual differences on the relationship between task complexity and interaction-driven learning opportunities. The results showed that as the task increased its complexity, learners’ language production 18 improved in terms of lexical diversity and accuracy while syntactic complexity appeared to decrease. With respect to how task complexity affects interactional modifications, it was found that increasing task complexity would lead to more interactional modifications, especially LREs. However, no clear effect of task complexity on other types of corrective feedback was detected. Another aspect of tasks that should not be ignored in interaction research is planning time. As Ortega (2005) pointed out, opportunities to think about language use and strategies to complete a task during planning time enable learners to produce a more accurate and complex language with higher fluency during actual task performance. From the cognitive perspective, during planning time, learners are able to draw upon their prior knowledge which allows them to allocate their attention and memory to other aspects of the tasks, making their performance more complex (P. Robinson, 2003). Similarly, Ortega (1999) argued that planning time mitigates cognitive load and communication pressures, which enables learners to perform the task better. Empirical research also lends support to the provision of planning time for its positive impact on the fluency and complexity of learners’ language production, although studies on planning time and accuracy show some mixed results (N. C. Ellis, 2009; Foster & Skehan, 2009; Ortega, 1999; Yuan & R. Ellis, 2003). In addition to planning time, another task variable that is worth considering when examining the relationship between tasks and L2 learning is task familiarity. To date, research studies have identified four key influential factors on the provision of L2 learning opportunities during task-based interaction. They are interpersonal familiarity, 19 content familiarity, procedural familiarity, and cognitive load familiarity. One of the earliest research studies which tap into the effects of task familiarity on learner production is Gass, Mackey, Alvarez-Torres, and Fernandez-Garcia (1999). In this study, learners of Spanish as a L2 were grouped into two experimental groups and one control group. The first experimental group watched a silent film clip three times, and were asked to narrate the film in the target language before they watched another silent film clip and described it. The second group watched four different silent film clips, and were asked to narrate each of them, while the control group only watched two of the four film clips. Learner production was examined for the accuracy of the Spanish verb ser and estar and lexical complexity, and the results showed that the first group outperformed the other two groups, which suggested that increasing task familiarity in terms of content would improve learner production. Along the same line, Mackey, Kanganas, and Oliver (2007) explored how task familiarity affects the amount and type of feedback, opportunities for modified output, and actual modified output produced by child L2 learners. Forty child ESL learners participated in tasks that vary in terms of content familiarity and procedural familiarity. Interestingly, the findings showed that unfamiliarity with task content and task procedure led to more interactional feedback such as clarification requests, confirmation checks, and corrective feedback. However, more actual modified output and attention to form were found in tasks that were familiar procedurally and content-wise. In another study in an EFL context, Qiu and Lo (2017) found that content familiarity and task repetition had a positive impact on Chinese EFL learners’ engagement in L2 use. In this study, sixty 20 EFL learners were asked to carry out four narrative tasks with two familiar topics and two different topics while still working in the same dyads. Data analysis from learners’ production and stimulated recalls showed that learners were highly engaged in tasks with familiar topics as they felt more confident and relaxed during task performance. With respect to L2 learning opportunities, fewer self-repairs were found when learners performed tasks with unfamiliar topics. However, little has been known about how integrated and independent speaking tasks affect the way learners interact with each other, and whether such interaction occurrence is beneficial for L2 development. According to Butler, Eignor, Jones, McNamara, and Suomi (2000), an independent speaking task refers to a speaking task that is based on stand-alone visuals or statements. In an independent speaking task, learners are required to rely on their own personal experience or general knowledge to complete the task. On the other hand, an integrated speaking task requires learners to use multiple language skills to complete the task (Butler et al., 2000). In integrated tasks, learners are either provided with a listening or reading input on which learners base their speaking performance. This type of task is argued to provide a favorable L2 learning condition because it simulates a real-life communicative need when learners have to use two or more skills to complete a task. However, the extent to which a task is successful in engaging learners and generating opportunities for L2 learning also depends on learners themselves. Therefore, it would be incomplete to claim the effects of interaction on L2 learning without taking the mediating effects of learner-related factors into consideration. 21 This leads to the next section which closely examines how interaction-driven learning varies in relation to learner-related factors. Learner proficiency and interaction-driven L2 learning. One of the factors that affect the provision of interactional modifications is learner proficiency. Recently, research studies have indicated that the amount, type and outcome of interactional learning opportunities are closely tied to the proficiency of dyadic members who engage in interactive activities. For example, Williams (2001) conducted a study in an intact learning environment to examine the occurrence and resolution of LREs in collaborative dialogues. Eight L2 learners of English of four proficiency levels (i.e., beginner, low intermediate, high intermediate, and advanced levels) participated in the study. She found that although the instances when learners focused on form while performing meaningfocused interaction were not high, it is evidenced that they were able to direct their attention to formal features. This is supported by previous researchers who found that lower proficiency learners seemed to place more emphasis on processing meaning than they did with form, suggesting that they were not developmentally ready to direct their attention to formal aspects during task performance (Farrokhi, Ansarin, & Mohammadnia, 2008; VanPatten, 1990, 1996, 2003; Williams, 1999). Further analysis of the findings showed that the probability that more competent learners resolved linguistic issues successfully was significantly higher than that of their lower level counterparts. Regarding the type of LREs occurring during interaction, the number of episodes in which learners focused on lexical aspects of the target language was far higher than that of grammar-related episodes. 22 Building on what Williams (2001) found, Leeser (2004) examined the oral performances of twenty-one dyads of L2 adult Spanish learners who were enrolled in a content-based course in order to identify the role of learner proficiency in the promotion of language learning opportunities. Learners were arranged into three types of dyads according to their proficiency levels (e.g., high-high, low-low, and high-low) and completed a dictogloss task. The research findings pointed out that learner proficiency displayed clear effects on the amount, type, and outcome of LREs produced. Like Williams (2001), Leeser’s study (2004) showed that dyads with two high proficiency learners produced more LREs and successfully resolved their linguistic issues than the other two types of dyads. Dyads with two lower proficiency learners produced the lowest number of LREs, and most of the LREs produced by this type of dyad were related to lexis. In contrast, dyads with two high proficiency learners showed a tendency to focus more on grammar. Drawing upon the observed results, it is hypothesized that lower proficient learners were not developmentally ready to discuss linguistic problems occurring during meaning-focused activities, which probably accounted for the high number of LREs left unresolved. Having explored the mediating effects of learner proficiency on the provision of interactional features that hold beneficial value to SLA, recent research on interaction has witnessed an increasing number of studies which explore the interrelationships between task types, learner proficiency, and L2 development. One of the recent research studies following this direction is Watanabe and Swain (2007). In this study, four adult intermediate ESL learners were paired with four higher proficient learners and four lower 23 proficient interlocutors. Learners participated in a three-stage task consisting of pair writing, noticing, and individual text reconstruction. Learners’ collaborative dialogues during these task stages were analyzed for LREs and patterns of pair interaction. The findings indicated that learners benefited from pairing with both lower and higher proficiency interlocutors but the benefits varied depending upon the stage of the task and the nature of the interaction. For instance, when paired with a higher proficiency interlocutor, learners tended to produce more LREs than when they collaborated with a lower proficiency learner. The post-test scores, however, showed that participants learned more when working with lower proficiency learners. Speaking of interactional patterns, it is found that when learners collaboratively performed the tasks, they generated the most number of LREs, irrespective of the proficiency of their interlocutors. Regarding the occurrences of LREs among different stages of the task, the research results revealed more LREs were significantly produced when learners worked with a higher proficient learner during the pair writing stage of the task. However, there were more LREs generated during the noticing stage when learners were paired with lower proficiency learners. From the finding of the study, we might assume that it is patterns of interaction, not learner proficiency, that have influenced the number of LREs occurring during collaborative tasks. In the same vein, Kim (2009) investigated the mediating effects of task complexity on the occurrence and resolution of LREs in dyadic settings. Thirty-four ESL learners with two different proficiency levels were assigned to perform two tasks (i.e., picture narrative and picture difference tasks) of two different complexity levels. The 24 results indicated that task complexity displayed its impact on the occurrence of L2 learning opportunities, and the impacts varied according to task types and learner proficiency. A closer analysis of learner interaction during task performance revealed a pattern that lower proficiency learners tended to produce more LREs in the simple version of the narration task while learners who were at the higher proficiency level produced more LREs during the complex version of the narration task. There were significantly more LREs generated by less fluent learners in a complex version of the picture difference task, whereas no significant difference in the number of LREs was found among higher proficient dyads. Regarding the resolution of LREs, it was found that learners were able to resolve more linguistic problems in complex tasks than in simple ones, irrespective of task types. These findings lend support to P. Robinson’s Cognition Hypothesis (2001), which claims that increasing task complexity in terms of resource-directing dimensions, is likely to facilitate L2 learning. Kim’s study goes a further step, contributing to P. Robinson’s Cognition Hypothesis by adding that task types and learner proficiency levels play a role in modulating the impact of task complexity on L2 learning and development. In addition to a number of interaction research studies in ESL contexts, efforts have also been made to provide more insights into the effects of learner internal factors on the interaction-driven learning opportunities in English as a Foreign Language (EFL) contexts. For instance, Kim and McDonough’s study (2008) investigated collaborative dialogues of Korean learners of English during two dictogloss tasks. In the study, intermediate learners were paired with other intermediate learners and then with other 25 advanced learners to reconstruct two dictogloss tasks. Learning opportunities were operationalized as the occurrence and resolution of LREs. It was found that when interacting with more advanced interlocutors, Korean L2 learners produced significantly more LREs than when collaborating with other intermediate partners. However, unlike other studies on LREs which found that learner proficiency appeared to have a mediating effect on the occurrence of lexical and grammatical language episodes during meaningfocused activities, Kim and McDonough (2008) found that there was no significant difference in terms of lexical and grammatical discussion when learners were paired with interlocutors of different proficiency levels. Analysis of the resolution of LREs also revealed that L2 learners seemed to gain more benefit in collaborative interaction with a more fluent learner as more resolved LREs were found in mixed proficiency dyads. Another issue worth considering is that learner proficiency also influences pair dynamics in collaborative tasks. In other words, when learners collaborate with interlocutors of different proficiency levels, interactional patterns change accordingly. According to Storch (2001), when learners work in dyads, their interaction follows one of the following interactional patterns: (1) dominant/dominant pair, (2) dominant/passive pair, (3) expert/novice pair, and (4) collaborative pair. In dominant/dominant interactional pattern, learners individually present their ideas without building on their interlocutor’s contribution, making it difficult to reach a consensus when disagreements arise. In the dominant/passive pair, a more proficient learner takes control of the interaction, leaving the other interlocutor little chance to contribute his or her own ideas in the conversation. The third interactional pattern (i.e., expert/novice pair) involves a 26 more knowledgeable learner working with a less proficient learner, and the more fluent learner acts as a resource or a facilitator, who provides linguistic assistance for the less proficient learner and leads the lower level learner to go beyond his or her current developmental zone. The last type of interactional pattern occurs when two members of a dyad contribute their ideas collaboratively by building on and developing their interlocutors’ ideas. Of the four types of pair dynamics, collaborative pair and expert/novice pair are claimed to be conducive to L2 learning. Following Storch’s interactional model, a large number of research studies have been conducted to examine whether Storch’s claim remains true in other educational contexts. Watanabe and Swain (2007) found that only the collaborative pattern of interaction was beneficial to L2 learning while other types of interactional patterns provided little opportunities for L2 development. They also found that in expert/novice pairs, only expert learners seemed to gain benefit from interaction through teaching their peers while novice learners failed to internalize information provided by expert interlocutors. One possible reason for this is that lower proficient learners are not developmentally ready for the uptake of new knowledge (Leeser, 2004). Another reason could be attributed to the high affective filter when collaborating with a more knowledgeable interlocutor, who might have limited the chance for less proficient learners to actively participate in language-related discussion and solve linguistic problems on their own (Swain & Miccoli, 1994). However, the research findings indicated that less proficient learners displayed some noticing benefits when they were 27 guided by the more advanced interlocutors (Kim, 2009; Kim, 2012; Leeser, 2004; Nuevo, 2006; Storch & Aldosari, 2013; Watanabe & Swain, 2007). It is also interesting to know that pair dynamics change when learners change their dyadic members. For example, L2 learners who seemed to adopt a passive role when paired with a more proficient interlocutor tended to play a collaborative role when collaborating with a less fluent interlocutor (Kim & McDonough, 2008). Although it is not conclusive that a mixed proficient dyad seemed to create a better condition for L2 learning than a dyad consisting of two learners of the same proficiency level, research studies have shown that pairing two learners of different proficiency levels is conducive to L2 learning only when the more proficient learner of a dyad takes an expert role rather than a dominating role during interactive tasks. That is to say, when the more advanced learners act as a facilitator during peer-peer interaction, they are more likely to use their knowledge while encouraging their less-proficient interlocutors to contribute their ideas and collaboratively perform the task. Insights from these research findings provide two important pedagogical implications in L2 classrooms regarding how to group students in pair work. First, when the dominant/dominant interactional pattern becomes prevalent in L2 classroom, language teachers should encourage learners to work with different partners to change pair dynamics. Second, it is necessary for language teachers to explain the importance of collaborative learning and direct learners about their roles when they work with other learners in order to maximize the benefits of interaction on L2 development. In summary, the previous studies on learner proficiency and interactiondriven L2 learning show that the proficiency of dyadic partners seems to be an important 28 factor that has an influence on the opportunities of interactional learning. Therefore, the current study looks at the effect of interlocutors’ proficiency on interactional patterns in different task types. Learners’ perceptions and interaction-driven L2 learning. Investigation of learners’ perceptions of the use of meaning-oriented tasks and peer-peer interaction is crucial, since information about learners’ mental processes could be used to supplement what researchers claim about the potential benefits of interaction through observable data, thereby increasing the validity of the interaction approach. Moreover, insights into learners’ perceptions about their interaction experience are important to understand the underlying factors that influence learners’ behaviors during interactive tasks. Two common ways that have been used to elicit learner perception regarding their experience and learning opportunities that happen during interaction are introspective methods (e.g., think-aloud, questionnaires, stimulated recall interviews, and diary-keeping) and verbal reports (e.g., concurrent and retrospective). Using these two methods, research studies to date have provided empirical evidence illustrating that there is a correlation between learners’ perceptions and their interactive task performance, and even their subsequent L2 development. For example, Dörnyei and Kormos (2000) indicated that learners were more likely to interact with their interlocutors and provide interactive feedback when they had a positive view towards the task. Schulz (2001) also added that learners’ expectations might also account for their engagement during tasks and their willingness to provide interactional modifications. In addition to that, it is claimed that learners’ experiences, their beliefs about 29 communication, and their beliefs about their interlocutors play an important role in shaping the nature of interaction (Lantolf, 2000). For instance, how learners perceive their interlocutors’ motivation during task performance affects their willingness to communicate and engage in the language learning process (Dörnyei & Kormos, 2002; Mackey, 1999). Along the same line, Mackey (2002) explored how learners perceived their role during interactional processes. The research study was conducted with forty-six ESL learners who were videotaped during three communicative tasks with their peers, their teacher, and native speakers in an intact classroom context. Learners’ recalled comments showed that the majority of episodes, which were identified by the researcher as L2 learning opportunities, were also acknowledged by learners as occasions for learning where input was made comprehensible through negotiated interaction. Similarly, a recent research study conducted by Watanabe and Swain (2008) explored how learners perceived their interlocutor’s proficiency and its impacts on the nature of the interaction. Forty-six ESL learners carried out a three-stage task in collaboration with higher and lower proficiency learners, and data from stimulated recall interviews showed that how learners perceived their partners’ proficiency level could determine the way they interacted and provided assistance during interaction. It is also evident in Philp and Mackey (2010) that learners are more likely to experiment with the target language, and get ready to attend to the provided corrective feedback with their trusted friends. In addition to that, there is an increasing recognition of the importance of social context in interaction research, with a great deal of research suggesting that contextual 30 factors and interpersonal relationships can have an impact on learners’ willingness to produce output, provide feedback and attend to form (Batstone, 2011; Philp & Duchesne, 2008; Tarone & Bigelow, 2005). As indicated in Philp and Mackey’s study (2010), learners were more willing to take linguistic risks when they worked in pairs and small groups with their peers. Analyses of learners’ language production also revealed that learners were more likely to engage in interactional processes and produced more modifications in peer-peer interaction compared to teacher-student interaction. It is also interesting that the extent to which learners trusted their interlocutors influenced the way they offered feedback and attended to feedback. While some learners only provided and attended to corrective feedback when they interacted with their trusted friends, others expressed the fear of threatening the relationship, which may limit their provision of negative feedback. Recently, researchers have pointed out the relationship between interlocutor familiarity and L2 acquisition. It is argued that when learners are familiar with their interlocutors in terms of how long they know each other, they are more likely to use signals of non-understanding and produce more interactional feedback (Cao & Philp, 2006; Plough & Gass, 1993; Poteau, 2011; Shirvan & Talebzadeh, 2017). Leeser (2004) also found that while learners tended to be more patient with unfamiliar interlocutors, they seemed to get more involved in task performance, and to produce more modified output and negotiation through interactional features, such as confirmation checks and clarification requests when interlocutor familiarity existed. As demonstrated by recent research studies, learners appear to have a sense of security, and feel less anxious, which 31 in turn leads to a better task performance when they work with a familiar interlocutor (O’Sullivan, 2002; Shirvan & Talebzadeh, 2017). Pre-test and post-test results of a semester-length study by Poteau (2011) indicated that learners were able to retain more targeted vocabulary when they worked with familiar interlocutors than when they collaborated with unfamiliar partners. In short, the findings of these studies suggest that learners’ perceptions towards tasks and interlocutors play a crucial role in their production of interactional patterns. Despite an increasing recognition about the importance of learners’ perceptions of classroom interaction and its impacts on interaction and L2 learning, little is known as to what learners think about interaction and its value to their L2 development. It is argued that without insiders’ perspectives about their experience during interactive tasks, the picture of interaction and L2 development would be incomplete. In summary, this chapter has reviewed research from the last two decades by describing recent interaction research studies on how task and learner characteristics interact with each other and jointly shape interaction-driven L2 development. The studies suggest that task characteristics, learner proficiency, and learners’ perceptions towards tasks are influential factors on interaction-driven learning. Moreover, a brief review of the literature shows that interaction research has expanded considerably over the last several decades that it is not sufficient to ask whether interaction impacts learning (Mackey, 2013), but how interaction impacts L2 development. Despite a growing number of interaction research studies which significantly move the field forward, there is still a continual need to have more research into the interrelationships among contextual factors, 32 task characteristics, learner internal variables, and interaction-driven L2 learning opportunities. Taken all together, the current research study is designed to investigate how different task types affect L2 learning opportunities during task-based interaction. Additionally, the study examines how learners’ interactional performance differ across different proficiency levels. Taking the importance of learners’ perceptions of their taskbased interaction into consideration, the current research study was also designed in a way that allows learners to freely express their thoughts and feelings about their interaction, so that their insights can be used to provide pedagogical implications to facilitate L2 learning opportunities and L2 development. Research Questions With the insights drawn from the previous research studies discussed in this chapter, this study was designed to seek the answers to the following questions: 1. How do the integrated speaking task and the independent speaking task differ in relation to the participants’ interactional performance? 2. How does the participants’ interactional performance differ across different proficiency levels? 3. What are the participants’ perceptions of the integrated speaking task and the independent speaking tasks? 33 Chapter III Methodology In order to answer the three research questions and better understand the impact of task types, learners’ proficiency levels and learners’ perception on the provision of interactional feedback during interactive tasks, a combination of qualitative and quantitative approaches was employed. A quantitative approach might enable the researcher to examine the relationships between task types, learner proficiency, and language learning opportunities, whereas through qualitative data collection procedures, detailed information on the participants’ perceptions and sociocultural factors that impact interaction-driven learning can be gathered. In what follows next, the research setting is described, followed by the information of the study participants, and the materials used for data elicitation. Finally, a description of data collection procedures and data analysis is presented. Research Setting The research was conducted with two groups of international students who were taking L2 courses at a U.S. Midwestern university during the Fall 2017 semester. The classes chosen for this study were English for Academic Purposes (EAP) 125 – Introduction to oral communication for multilingual speakers, and Intensive English Program (IEP) 022 – Low-intermediate Listening and Speaking. The two classes are designed to prepare L2 learners to enter university coursework, focusing particularly on developing learners’ listening and speaking skills. According to the general description of the course from the university catalog, the primary goal of EAP 125 is to develop 34 students’ academic listening and speaking skills to prepare them for their academic studies. These skills include but are not limited to listening to authentic academic lectures, taking notes, participating in small group discussions, study skills, and giving oral presentations. On the other hand, IEP 022 is set to develop listening and speaking skills for students at the low-intermediate level by involving them in a variety of academic and social conversations, short oral presentations, and a variety of academic lectures. Participants There were twenty-four international students who voluntarily participated in this current research study, ten of whom came from EAP 125, and fourteen of whom came from IEP 022. Since the data were collected over two different class periods, only eight participants attended both sessions, four of whom came from IEP 022, and the remaining four came from EAP 125. Only data from those eight participants were used for further analysis. Pseudonyms were used for all participants in the study for privacy. Regarding the demographic information of the chosen participants, they were all F1 holding visa students, coming to the U.S. in pursuing a degree program. They were varied in majors and had various L1 backgrounds: Japan (3), South Korea (1), Indonesia (1), India (1), Ivory Coast (1), and Kuwait (1). Two of the participants were female and the other six were male. Their age ranged from 18 to 26, with a mean age of 20.25 years. At the time of data collection, all participants had been living in the US for an average of four months, with an average of six years of previous formal English instruction. At the university where the research was conducted, students were placed into classes based on 35 their proficiency level measured by standardized tests (e.g., IELTS or TOEFL iBT) and/or the placement test provided by the university. According to the university placement policy and the biographic information provided by the participants, learners enrolled in EAP 125 were considered advanced learners whose average TOEFL iBT score is around 80. Learners from IEP 022 were at the low intermediate level who had an average score of 30 on the TOEFL iBT test. Materials In order to elicit data to answer the research questions, the following instruments consisting of two oral communicative tasks (i.e., an integrated speaking task and an independent speaking task) and two questionnaires were developed. Tasks. Overall, the two tasks used in this study are two-way in that they require the participants to exchange information for a successful task completion. The tasks are also convergent in the sense that the participants have to negotiate to reach an agreement on a discussed issue. However, what sets the two tasks apart is that the integrated task requires the participants to integrate what they heard from the video input into their discussion whereas there is no such input provided in the independent task. Following is the detailed description of each task. For the integrated speaking task, two videos were created, in which two native speakers from North America were invited to talk about their opinions about social media. The speakers were instructed to talk about either the advantages or the disadvantages of social media. Each video clip lasts approximately two minutes. To assist the participants in the task, step-by-step instructions were given to the participants before 36 they began the task (see Appendix A). In detail, the participants were asked to watch their given video and summarize what they heard to their partner. Then, they were asked to negotiate and reach a consensus on whether they thought social media has a more positive or negative impact on people’s lives. In order to complete the task, participants had to give at least three reasons for their decision. For the independent speaking task, participants were provided with a written prompt “Should students work while attending college?”. Then, they were asked to individually list the pros and cons of the issue before coming together in pairs to discuss and come to an agreement on whether they were both more in favor of or against students’ having jobs while they were studying at college. Like the integrated task, the participants had to give at least three reasons for their decision. A task description was also given to the participants to assist them in completing the task (see Appendix B). Questionnaires. In order to elicit learners’ perceptions of their task-based interaction, two post-task questionnaires consisting of both close-ended and open-ended items were developed (see Appendix C and D). The close-ended items target what participants thought about the tasks in terms of task interest, task difficulty, task familiarity, task appropriateness, and task helpfulness. Each item was designed with a 5point Likert scale, ranging from strongly disagree, disagree, neutral, agree, to strongly agree, with values corresponding from one to five. The second part of the questionnaire consists of several open-ended questions, allowing participants to further elaborate about their experience during task-based interactions. Demographic information of the participants, such as age, gender, L1 background, years of studying English and time 37 spent in the U.S., is also included in the post-task questionnaire. To ensure instrument validity, all of these instruments were checked by an experienced professor of English and were piloted with two advanced students who did not participate in the current study. Data Collection Procedure The study was carried out during regular class time in the Fall 2017 semester. The data collection process was conducted separately in EAP 125 and IEP 022, but the procedures remained the same for the two classes. First, the researcher visited the classes with the permission of the instructors, introduced the purpose of the study, and answered any questions that students might have before administering the consent forms. At that point, students had the choice to participate in the study or not. After consent forms were collected, participants were organized in pairs and instructed on the procedure to perform the first task (i.e., an integrated speaking task). In the first task, each participant of each dyad was provided with an opinion-based video about social media. One participant of each dyad was assigned to watch a video which presents the speaker’s opinion about the advantages of social media, while the other participant watched a different video about the disadvantages of social media. After watching the video twice, each participant was given five minutes to summarize for their interlocutor what they had understood from the video. Then, they were asked to discuss their viewpoint about the topic and come to consensus whether they thought social media had a more positive or negative impact on people’s lives. The participants were given 20 minutes to do the task, and the whole discussion was audio-recorded for further analysis. Following the discussion, a questionnaire (see Appendix C) was administered to elicit the 38 participants’ perception of the task and their peer interaction. The participants were asked to do the second task on another day. In the second task (i.e., an independent speaking task), participants worked in the same pairs. They were provided with a task description and a written prompt “Should students work while attending college?”. They were asked to individually list the pros and cons of the issue before coming together, discussing and coming to consensus whether they were both more in favor of or against students’ having jobs while studying at college. Participants were also given 20 minutes to do this task. After finishing the task, they were asked to answer another questionnaire (see Appendix D) which was almost the same as the first questionnaire. Like the first task, the participants’ speaking performance was audio recorded for subsequent data analysis. It should be noted that except for the clarification of the task procedures, there was minimum intervention from the researcher during the learner interactions. Data Analysis To begin the process of analyzing data, the audio recordings of the learner interactions were transcribed following transcription conventions adapted from Poupore (2004) (see Appendix E). The transcriptions were then coded for six interaction features: negotiation of meaning, negotiation of form, negotiation of task content, negotiation of task procedure, negotiation of personal experience, and self-initiated repair. Following are the definitions of the six focal interactional features adopted from previous research studies (Adams & Ross-Feldman, 2008; Poupore, 2004). 39 Negotiation of meaning (NM): “side sequences to the main flow of conversation aimed at signaling and solving problems of message comprehensibility, that is aimed at restoring mutual understanding” (Poupore, 2004, p. 243). Negotiation of form (NF): “side sequences to the main flow of conversation aimed at drawing the interlocutor’s attention to formal aspects of what was produced and encouraging ‘self-repair’, or at the very least, acknowledgement of the formal modifications that the listener suggested” (Poupore, 2004, p. 243). Negotiation of task content (NC): “stretches of interaction aimed at pushing the interlocutor to provide more information than spontaneously offered in relation to task content” (Poupore, 2004, p. 244). Negotiation of task procedure (NP): “stretches of interaction aimed at pushing the interlocutor to provide more information than spontaneously offered in relation to task procedure” (Poupore, 2004, p. 244). Negotiation of personal experience (NPE): “stretches of interaction aimed at pushing the interlocutor to provide more information than spontaneously offered in relation to personal experiences” (Poupore, 2004, p. 244). Self-initiated repair (SIR): any instance when a learner modified his or her own utterance in the turn or in an adjacent turn without indicators from his/her interlocutor (Adams & Ross-Feldman, 2008). As discussed in Chapter 2, not all negotiation moves result in L2 learning. Therefore, to better understand the effectiveness of these tasks on the learning of these participants, the researcher calculated the number of successful (quality) interactions. 40 Quality interactions in this study are considered to be responses in which learners are able to: (1) modify their output to be more comprehensible and/or more target-like in terms of grammatical, phonological, or morphosyntactical aspects, or (2) provide additional information in response to interlocutor’s feedback on the incomprehensibility or incompletion of the original utterance (Pica et al., 1996; Poupore, 2004; Shehadeh, 1999; Van den Branden, 1997). In other words, in order for an interaction to be considered as a quality interaction, the utterance following the indicator of incomplete comprehension must be more accurate, more comprehensible or more informative addressing the concern raised in the indicator. The quality interactions were also categorized into six different interactional feature groups. To address the first research question (i.e., How might the integrated speaking task and the independent speaking task differ in relation to the participants’ interactional performance?), the six focal interactional features (i.e., negotiation of meaning, negotiation of form, negotiation of task content, negotiation of task procedure, negotiation of personal experience, and self-initiated repair) were coded for the frequency of their occurrences. Then, the percentage of successful (quality) interaction (i.e., interactional moves that lead to L2 learning opportunities) was calculated. The distribution of the six focal interactional features was displayed according to the two interactive tasks to highlight how the tasks differ in terms of interactional features produced. Drawing on the initial data process, the distribution of the six focal interactional features, including the number of interactional moves and the percentage of successful 41 interaction, was presented in accordance with the two groups of learners (e.g., advanced learners in EAP 125 and low-intermediate learners in IEP 022). This might enable the researcher to explore how learner proficiency levels mediated the emergence of interaction-driven learning opportunities, providing an answer to the second research question (i.e., How might the participants’ interactional performance differ across different proficiency levels?). To answer the third research question (i.e., What are the participants’ perceptions of the integrated speaking task and the independent speaking task?), the participants’ responses to close-ended items in the post-task questionnaires were put into the statistical software program SPSS (version 19.0) for descriptive and inferential statistics. For inferential statistics, the dependent sample t-test was applied, with alpha set at 0.05 to determine statistical significance. The participants’ responses to open-ended questions in the questionnaires were analyzed qualitatively with the support of a software program called Nvivo (version 11.0) to identify emerging patterns or themes regarding the students’ perceptions of their task-based interactions. To ensure interrater reliability, a graduate student was recruited for coding data. After the norming session provided by the researcher, 10% of the data was coded by the two raters, with the agreement rate reaching 68%. Disagreements were then resolved through discussion between the two raters until total consensus was reached. The researcher then coded the rest of the data in considering what had been discussed with the second rater. Following the data analysis process, the main findings of the research are revealed and presented in relation to each of the three research questions in the next chapter. 42 Chapter IV Results How might the Integrated Speaking Task and the Independent Speaking Task Differ in Relation to the Participants’ Interactional Performance? The following table (Table 1) represents the distribution of the six interactional features and the percentage of quality interaction by each interactional feature according to the integrated speaking task and the independent speaking task. Table 1 Interaction variables across tasks Integrated speaking task Independent speaking task q pq n pn q pq n pn NC 10 25% 8 80% 8 18% 5 63% NM 8 20% 3 38% 15 34% 7 47% NF 5 12.5% 5 100% 0 0% 0 0% NP 13 32.5% 3 23% 16 36% 4 25% NPE 4 10% 2 50% 5 12% 3 60% Total Ns 40 100% 20 50% 44 100% 19 43% SIR 12 12 11 11 Note. NC: Negotiation of Task Content, NM: Negotiation of Meaning, NF: Negotiation of Form, NP: Negotiation of Task Procedure, NPE: Negotiation of Personal Experience, SIR: Self-Initiated Repair, n: number of negotiation moves, q: number of quality negotiations, pn: percentage of negotiation, pq: percentage of quality negotiation Overall, the independent speaking task produced slightly more interactional modifications than the integrated speaking task, with 44 and 40 negotiation moves respectively. However, more quality negotiation was found in the integrated speaking task than in the independent task, with the percentage of quality interaction accounting for 50% and 43 % in turn, although the difference was not significant. 43 A closer look at the data reveals how interactional features differ across the two tasks. The most striking difference between the two tasks lies in negotiation of meaning, with the amount of negotiation of meaning produced in the independent task (15) nearly doubling that of the integrated task (8). The percentage of quality negotiation produced by this interaction feature was relatively high, 38% for the integrated task and 47% for the independent one. It is also interesting to find that no negotiation of form was detected in the independent speaking task, whereas in the integrated task, the learners negotiated on form five times. The integrated speaking task appeared to produce more quantity and quality negotiation of content while more quantity and quality negotiation of meaning, negotiation of procedure, and negotiation of personal experience were observed in the independent speaking task. However, these differences were insignificant. Regarding self-initiated repairs, the difference between the two tasks was insignificant, with 11 occurrences for the independent task and 12 for the integrated task. In addition to the differences mentioned above, there were some common patterns in the distribution of interactional features across the two tasks. It is interesting to note that, the most common type of interaction in both tasks was negotiation of procedure, constituting one-third of the total negotiation moves of each task (32.5% of the integrated speaking task and 36% of the independent speaking task). Despite its highest frequency, the percentage of quality interaction produced by negotiation of procedure was the lowest, around 23-25% respectively. On the contrary, negotiation of personal experience only made up a small proportion, around 10-12% of the total negotiation moves observed. However, the percentage of quality interaction produced by this type of 44 interactional feature was the second highest, with the successful ratio reaching 50% in the integrated speaking task and 60% in the independent task. In addition, despite the variation in the frequency of negotiation moves across the two tasks, it is found that the percentage of quality interaction generated by negotiation of content was always higher than that of other negotiation moves, with 80% in the integrated task and 63% in the independent task. Among the focal negotiation moves, negotiation of form was the least frequent interaction feature across the two tasks. How might the Participants’ Interactional Performance Differ across Different Proficiency Levels? With regard to the mediating effects of proficiency levels on interaction-driven learning opportunities, Table 2 highlights the differences in terms of how the two groups of learners gave interaction feedback during their task-based interactions. In general, it was found that the advanced group seemed to produce more quantity and quality interaction than the lower group in most of the focal interactional features. In total, there were 46 negotiation moves generated by the advanced group, compared to 38 for the lowintermediate group. Similarly, the percentage of successful negotiation produced by the advanced learners was 54%, which was 17% higher than that produced by the lowintermediate learners. Another noticeable difference between the two groups of learners is that the advanced learners were able to repair their own utterance more frequently than their lower-proficiency counterparts, with the number of self-initiated repairs produced by the advanced group (15) nearly doubling that of the lower group (8). 45 Table 2 Interaction variables across proficiency levels EAP (n=4) IEP (n=4) (Advanced level) (Low-intermediate level) n pn q pq n pn q pq NC 12 26% 10 83% 6 16% 3 50% NM 12 26% 7 58% 11 29% 3 27% NF 3 6.5% 3 100% 2 5% 2 100% NP 16 35% 4 25% 13 34% 3 23% NPE 3 6.5% 2 66% 6 16% 3 50% Total Ns 46 100% 25 54% 38 100% 14 37% SIR 15 8 Note. NC: Negotiation of Task Content, NM: Negotiation of Meaning, NF: Negotiation of Form, NP: Negotiation of Task Procedure, NPE: Negotiation of Personal Experience, SIR: Self-Initiated Repair, n: number of negotiation moves, q: number of quality negotiations, pn: percentage of negotiation, pq: percentage of quality negotiation Regarding the other interactional features, we can see that the advanced learners tended to produce more quantity and quality negotiation of content, negotiation of meaning, and negotiation of procedure than the low-intermediate level learners. However, what makes the two groups different the most is the frequency and the percentage of quality interaction generated by negotiation of content and negotiation of meaning. Specifically, the advanced group negotiated for content twice as frequently as their lower-level counterparts (12 compared to 6). The percentage of successful negotiation of content by the advanced group was also considerably higher than the lower group, with 83% and 50% respectively. Similarly, despite producing approximately the same amount of negotiation of meaning with the lower group (with 12 and 11 times in turn), the advanced group was able to generate 58% of quality interaction, more than doubling that of the lower group (with only 27%). 46 As for other negotiation moves, no significant difference was detected across the two proficiency levels. For example, the lower group produced twice as much negotiation of personal experience as the advanced group (with 6 and 3 respectively), but the percentage of quality negotiation produced by the latter group was 16% higher than that of the former group. With respect to negotiation of procedure, in spite of being the most common negotiated feature in learner-learner interactions, accounting for 34-35% of the total negotiation moves produced by the low-intermediate group and advanced group respectively, the percentage of quality interaction remains the lowest, around 23-25%. Interestingly, this pattern holds true regardless of task types, as presented in the previous result section. It is also noticeable that the learners, irrespective of their proficiency levels, rarely negotiated on form, making up only 5-6% of the total interactional features. However, they were able to resolve all the form-focused episodes successfully. The following excerpt (1) displays a LRE in which a dyad of advanced proficiency was able to resolve a phonological problem during their interaction. In this excerpt, learner F did not pronounce the word ‘isolation’ correctly (line 3), resulting in learner E giving a clarification request (line 4) signaling that he did not fully understand her utterance. Initially, learner F might have thought that learner E did not hear the word ‘isolation’, so she repeated the word with the same incorrect pronunciation. Only after the second negative feedback from learner E that learner F realized the issue and modified the pronunciation of the word in a target way (line 7). 47 (1) 1 F: Uh, meanwhile I don't know what she is doing but she give three, she 2 give three points, three opinion about the negativity of social media. The 3 first one is that it, it can increase isolation { /ˌɪsəˈleɪʃən / }, like/ 4 E: increase? 5 F: 6 E: Relation? 7 F: 8 E: Isolation? Isolation, okay isolation. Isolation. { /ˌɪsəˈleɪʃən / } Isolation. { /ˌaɪsəˈleɪʃən / } Similarly, the following excerpt (2) presents how learner F mispronounced the word ‘vulnerable’, which hindered her message comprehensibility (line 1). This led learner E to give negative feedback by repeating part of the previous utterance with rising intonation, indicating that he did not get the word (line 3). In response to that clarification request, learner F not only modified her pronunciation of the word ‘vulnerable’ to be more target-like (line 4) but also provided more context to help learner E better understand her intended message (line 6-7). (2) 1 F: 2 And the third thing is that it’s, it’s not good because users are vulnerable. (/vʌlˈnerəbəl/) 3 E: Are? 4 F: Vulnerable (/ˈvʌlnərəbəl/) 5 E: Vulnerable, okay. 6 F: Vulnerable, yeah, vulnerable because because they can be subject of scams, 7 8 theft, hacking and all the stuff. E: Okay. 48 Not only did the participants resolve phonological issues during their peer-peer interactions, but they also focused their attention on grammar. The following example (3) illustrates how a dyad of the low-intermediate level was able to benefit from negotiated feedback focusing on yes/no question structure. As can be seen in the example, learner N missed the main verb ‘use’ in his yes/no question (line 3). Having recognized the issue in learner N’s utterance, learner M gave implicit negative feedback by repeating part of the incorrect sentence with a pause at the place where the main verb was missing (line 5), with the hope that his partner could recognize the issue. However, learner N seemed not to realize the signal from learner M and just repeated part of his previous speech. Finally, learner M had to explicitly provide the target form (line 7) and resolved the episode. (3) 1 N: Uhm, together decide whether social media has more of a positive or 2 negative impact on people’s lives. Give three reasons for your 3 decision. Social media has a lot of positive impact, I think. Do you 4 social media network? 5 M: Do you … social media? Do you social media network? 6 N: Social media, social media. 7 M: Do you use social media? 8 N: Use, use. 9 M: Ah, yeah, yeah, yeah. I got you. I usually use Line. In the same way, the example (4) shows how another dyad of the lowintermediate proficiency level benefited from peer-peer interactions. In this example, learner S produced a non-target like utterance, using a bare infinitive as a subject of a 49 declarative sentence (line 1). Realizing the grammatical issue in learner S’s utterance, learner T offered a recast by repeating part of the previous non-target like utterance in a target way (line 2). In response to learner T’s recast, learner S was able to modify her utterance to be more target-like (line 3). (4) 1 S: Yeah, but take care, to connect, take care is very important. 2 T: Taking care? 3 S: Taking care is very important. Apart from these form-focused episodes, evidence from the collected data also shows that the participants were able to produce pushed output by providing more information to back up their viewpoint given the clarification request from their interlocutor. In the following example (5), learner S expressed her negative attitude towards social media but failed to provide the reason why she thought the way she did (line 1). In encouraging learner S to further elaborate her view, learner T gave a clarification request (line 2), indicating that she would like to know why learner S had such a negative viewpoint towards social media. This pushed learner S to give additional information to her original utterance, making it more comprehensible to her interlocutor (line 3-4). (5) 1 S: I think social media is not good. 2 T: Why? 3 S: Yeah, especially for children, social media, yeah, it only gives 4 5 fake news. T: Yeah 50 Another example of pushed output is presented in the following excerpt (6), which illustrates how learner N and M (both were at the low-intermediate level) negotiated on task procedure where they had to give three reasons why they thought social media has more positive or negative impacts on people’s lives. At the beginning of the episode, learner N asked his partner for help with the three reasons, but the way he asked seemed to confuse learner M (line 1). This led learner M to give a comprehension check (line 2). In response to the comprehension check, learner N modified his question and made it clear that he wanted to ask what he had to write about social media (line 3), which enabled learner M to provide an appropriate and detailed answer (line 4-5). (6) 1 N: What should I write? 2 M: What should you write? 3 N: What should I write down about social media? 4 M: You could say social media is, social media can make friends and 5 relationship. Likewise, in this example (7), learner I and J (both were at the advanced level) were talking about the pros and cons of having a part-time job during college. Learner J, in response to the question raised by learner I that if he would like to take a part-time job (line 1-2), provided a short answer which was not detailed enough (line 3). This led learner I to give a clarification request (line 4), which successfully pushed learner J to provide more information by giving a personal reason why he would take a part-time job (line 5-7). 51 (7) 1 I: 2 Yeah, it’s really hard --- --- -- Are you going to like if if you have time for part time job, are you going to do it or? 3 J: Yeah, I’m going to do it. 4 I: Because? 5 J: Yeah, I’m gonna do it, yeah, because I’m a kind of, like, I like travelling 6 so I need some money, like, in addition to what my parents give me to 7 do those travels. Regarding self-initiated repairs, it can be found in the data that while producing utterances, the participants were able to recognize the issue in their speech and repair it in a target way, even without any specific feedback from their interlocutor. The following excerpt (8) illustrates how learner M (at the low-intermediate level), while expressing his opinion about the use of social media, missed copula ‘be’ which should have followed ‘have to’ (line 2). Although he did not receive any negative feedback from his interlocutor, he was able to repair his utterance in his following turn (line 4). (8) 1 M: 2 Social media, social media can increase isolation, I think, social media is good, good too but we have to careful. 3 N: Yeah. 4 M: We have to be careful. For example, if you post picture or, picture or 5 movie for Facebook or Twitter, this picture can see everyone in the 6 world. If you post dangerous picture, or sexual picture, if you post, 7 anyone injure hurt, some problem. 52 In another case (9), learner E (at the advanced level) successfully repaired his utterance within the same turn. In providing reasons why students should not work while they are attending college, learner E figured out that his sentence was not target-like and immediately restructured his utterance to be more accurate. The self-initiated repair is marked with the underlined sentences (line 3-5). (9) 1 E: Yes, because when you when you work, you know that there are 2 something that you can do, and there are some, you learn learn to do good 3 choice, you learn to do the good choice. The cons, other than that, while 4 studying, less time for study. When you work, you have less time for 5 studying. The second thing is that/ From the examples mentioned above, it is evident that peer-peer interactions can bring beneficial elements to L2 learning. Through interactional feedback and interactional modifications, the participants of this study became aware of the issues that they had in their interlanguage regarding phonology, syntax, and morphology while being able to modify their utterances to be more target-like and/or more comprehensible to their interlocutors. Even without any specific feedback from interlocutors, the participants also benefited from communication repairs during communicative tasks. What are the Participants’ Perceptions of the Integrated and Independent Speaking Tasks? Table 3 presents the descriptive statistics of the close-ended responses of the participants regarding how they perceived the tasks they performed. Although there was variation among the criteria (e.g., task interest, task difficulty, task familiarity, task 53 appropriateness, and task helpfulness), the difference was insignificant. As illustrated in the table, the participants seemed to find the integrated speaking task more interesting than the independent speaking task, with the mean score and standard deviation as follows: M=3.75, SD=.46 for the former task and M=3.63, SD=.74 for the latter task. The participants also generally thought that the integrated speaking task should be used more often in a speaking class and even in a placement test, as indicated with higher mean scores and lower standard deviation than those of the independent task. On the other hand, the independent speaking task was found to be more challenging (M=3.50, SD=1.20) compared to the integrated task (M=3.00, SD=1.07) even though the participants thought that they were more familiar with the independent speaking task compared to the integrated one, with M=3.50, SD=0.93 and M=3.38, SD=1.06 respectively. In terms of task direction and planning time, the participants expressed that the independent speaking task had a clearer direction and the time given to complete the independent task was also more adequate than the integrated task. Despite the differences in attitudes towards the use of the two tasks, the participants appeared to agree that these two tasks equally helped them demonstrate their speaking skill, with the mean score of 3.38, but less deviation was found in the independent speaking task (SD=0.93) compared to the integrated one (SD=1.41). In order to examine whether these differences are significant or not, the dependent sample t-test was used, which showed that none of these differences was statistically significant. 54 Table 3 Learner perceptions across tasks (n=8) Integrated speaking task M SD 1. Task interest 3.75 0.46 2. Task difficulty 3.00 1.07 3. Clear direction 4.05 0.76 4. Adequate time 3.75 1.39 5. Use in speaking class 4.00 0.53 6. Task similarity 3.38 1.06 1.41 7. Reflect speaking skill 3.38 8. Use in placement test 3.88 0.64 Note. n: number of participants in the study Independent speaking task M SD 3.63 0.74 3.50 1.20 4.13 0.83 4.00 0.93 3.75 0.71 3.50 0.93 3.38 0.92 3.38 0.74 In order to know how the participants perceived their task-based interactions, the participants’ open-ended responses were gathered and analyzed for emerging themes. All of the participants were involved in responding to the open-ended questions. The responses’ length ranged vastly from zero to 64 words, but most of the responses were rather short (fewer than 20 words). Two of the participants, who were at the lower level, skipped one or two questions which ask them for their suggestions to improve the tasks. The results showed that the participants generally had a positive experience with these interactive tasks. Most of the participants (seven out of eight) said that they enjoyed the tasks and appreciated the discussion part in which they “could express [their] own opinions and get to know [their] partners’ ideas, which in turn helped to expand [their] knowledge about the topic”. One participant remarked that, “sharing what I retained of my video with my partner was funny and interesting, especially when I discovered that he had the opposite argument which is against my main idea”. This highlights the fact that 55 jigsaw tasks in which each learner are given different piece of information can be a good start to build the learners’ interest and curiosity. Another participant reported that “I really like the agreement form, the fact of figuring out pros and cons, and the amount of time given. Also in the video, the speaker was clear and specific.” In commenting about the integrated task, another participant expressed that “the task was interesting because we were able to follow up some videos and talk about it. Since it was an interesting task, we were also tested our listening too. So overall it was good”. These positive remarks provided useful information about the characteristics of a speaking task that can engage the students. With respect to how the topic affects learner interactions, two participants stated that, “the tasks helped organize [their] speaking and make it direct”. One participant shared that these topics affected him in a positive way, allowing him to share his knowledge with his partner and express himself without any pressure. Another participant further supported that by saying, “We can speak with other people to share our thoughts and mind changing. I decided I never attack on social media”. She then continued commenting that, “This became an opportunity for me to think deeply and critically about taking a part-time job”. Another participant reported that, “actually it taught me to know a lot of words to add to my speaking ability. I learn new words through the tasks”. In general, both tasks seemed to be successful in triggering critical thoughts about the discussed topics while offering the participants a chance to learn topic-related vocabulary, which might help the students express themselves. 56 It is also clear from the participants’ responses that topic also played an important role in enhancing learner engagement in task performance. As shared by one participant, “I like the topic, social media is always a good idea of topics because there are so many things to say about it”, while another expressed that, “I was able to know positive points and negative points about social media, so it was interesting”. Another learner said that, “We talked about the topic that we can face in university. The topic was ‘part time job’, and it is a close problem for me as a university student”. All these comments from the participants suggest that social media and part-time jobs are good topics for language learning activities because they are familiar to students, which enables them to relate the topic to themselves and engage in the tasks. Although “it is a subject known by everyone” as one participant commented, the participants’ responses indicated that if it is designed in such an interactive way, it can promote learners’ engagement which is the key to potential learning opportunities. In addition to the positive perceptions of the two speaking tasks, the participants also pointed out several challenges associated with their task-based interactions. First, the results showed that using video input with natural speaking rate seemed to be a challenge for learners, particularly the lower level learners. The majority of them (three out of four lower level learners) expressed that it was difficult for them to understand the video content because of their limited vocabulary and the speaking rate of the speakers. As shared by one participant, “It was difficult for me to listen [to] English. My English skill is low, and the speaker speed is very fast. I sometimes can't listen to video. I can't understand some vocabulary, and it is difficult to say my summary”. Indeed, while all 57 advanced level learners in the study reported that they could understand more than 80% of the listening input, half of the participants in the low-intermediate level addressed that they could understand 40% to 60% of the video. The second common problem that the participants, especially those who were at the low-intermediate level, experienced during their task-based interactions is related to producing output. Three out of four lower-level learners expressed their frustration in not being able to say what they wanted to say clearly, commenting that “I wasn't able to explain well and fluently” or “I don't speak well as I thought”. Other participants remarked that, “Our speaking skill is poor that makes our conversation short” while another one had trouble in “finding pros and cons for the independent speaking task.” Another learner reported that “My vocabulary was poor so I couldn't speak English immediately. I think after watching this video, I think I should learn more”. Drawing on these learners’ remarks, it can be said that although learners seemed to struggle during their interaction, engaging in these interactive tasks helped them be aware of the gap in their current interlanguage which might drive them to work more to push forward their language development. Third, while most of the lower level learners expressed their difficulty in listening and understanding the content of the video input, one advanced level learner reported that what made him dislike the task is its familiarity, saying that “It was nothing new, I have already done this type of task many times, and to test international students, you should give a tough test, I guess!”. Other than that, one of the participants said that he was afraid of negotiating with their partner for a fear that “If we have a different opinion, we will be 58 fighting”. It is not clear as to the reason behind the participant’s unwillingness to show his different opinion, but it might be related to the cultural differences in interaction which is common among Asian countries such as Indonesia and Japan where social consensus is emphasized. When asked about how the participants would like to improve the tasks, several suggestions were provided. While two learners would like to “make the tasks more interactive and funny”, one participant suggested that “If there were more details, it was better to understand and speak about it more”. Another participant also commented that the time given was not enough and that there should be more time for them to prepare a better conversation. Interestingly, one participant gave the idea of putting subtitle into the video, so that they could understand it more clearly. In terms of task implementation, one learner recommended that, “We should watch the video together, so we can share opinions immediately”. He then further explained that his partner was not ready to speak when he had been already ready. This brings up an issue of activity management on part of the teacher and or researcher who is going to use video as learning materials in class. In summary, the chapter highlights the main findings of the research regarding the impact of external factors (e.g., task types) and internal factors (e.g., learners’ proficiency levels and learners’ perception of task-based interactions) in mediating the occurrences of interactional feedback. The research findings showed that there was no significant difference between the integrated task and the independent task in terms of providing interaction-driven learning opportunities, with the exception of negotiation of meaning and form. Moreover, it is noticeable that among the six interactional features, negotiation 59 of procedure was more prevalent than any other types of interaction, irrespective of the task types and the participants’ proficiency levels. However, the number of quality interactions produced by this interactional feature was the lowest, which calls for a further discussion in connection with the previous research findings in the literature. In terms of the relationship between proficiency levels and interaction-driven learning opportunities, the advanced learners were shown to negotiate on content and repair their own speech significantly more than their lower counterparts. Taken all together, how task types, together with the participants’ proficiency levels and their perception of the tasks may have contributed to the provision of interactional feedback and interactional modifications will be discussed in the following chapter. 60 Chapter V Discussion The aim of this study was to explore the impact of two tasks (i.e., an integrated speaking task and an independent speaking task) on interaction-driven L2 learning opportunities and how these may differ in relation to learner proficiency levels. In addition, learners’ perceptions of the use of the two tasks in a L2 classroom context were also examined. These issues are discussed respectively in this section. Interaction-driven Learning Opportunities across Interactive Tasks The first research question focused on how the integrated and the independent speaking tasks impact interaction-driven L2 learning opportunities, which were considered as quality interaction (i.e., modified output and self-initiated repairs) in this study. Drawing upon the research results, it is evident that the two interactive tasks were able to promote opportunities for L2 learning through task-based interactions. Although the percentage of quality interaction produced is not exclusively high (50% in the integrated task and 43% in the independent task), it is still promising as for nearly half of the time of interaction, the participants were able to direct their attention to form and resolve LREs successfully while also paying attention to meaning to complete the tasks at the same time. Given the cognitive load that the participants had during task completion, the potential of the tasks in terms of creating L2 learning opportunities should be acknowledged. To connect it with the literature, this finding echoes what has been found in previous interaction research that not all interactional feedback leads to actual 61 modified output or uptake as there may be insufficient engagement in interaction or learners may fail to attend to feedback (Fujii, Mackey, & Ziegler, 2011). A comparison of the two tasks regarding the provision of interaction-driven learning opportunities indicated that the two tasks seemed to be equally beneficial, with the quantity and quality of interactional modifications being comparable across the tasks. The integrated speaking task was found to produce slightly more quality interaction than the independent speaking task, although the latter task produced more negotiation moves than the former one. However, it should be highlighted that these observed differences were insignificant, which does not lend support to P. Robinson’s Cognition Hypothesis (2003). According to this hypothesis, tasks with the provision of ideas are considered less cognitively challenging than tasks that require learners to come up with their own ideas. Additionally, the hypothesis predicts that the more cognitive challenging a task is, the more interactional feedback it will generate. Based on this hypothesis, the independent speaking task is more cognitively demanding than the integrated speaking task as the later provides learners with initial ideas related to the topic in the video input, which could serve as a resource for student discussion. If the Cognition Hypothesis holds true, there would be significant more interactional processes in the independent speaking task compared to the integrated task, which seems not to be the case in this study. Although there were not many differences between the two tasks in terms of the distribution of negotiation features, it is interesting to find that the amount of negotiation of meaning produced in the independent speaking task doubled that of the integrated speaking task. This finding is in line with what has been found in previous research 62 studies that learners tend to produce more negotiation of meaning during the more complex task version (Gilabert, Baron, & Llanes, 2009; P. Robinson, 2011). Moreover, the fact that significantly more negotiation of meaning was observed in the independent speaking task may be attributed to task familiarity. As the independent task was introduced after the integrated task, the participants may have familiarized themselves with what they were expected to do. This might enable them to focus more on the content of the task and/or the language involved in the task, which supports the general belief that familiarity with the task itself leads to a greater amount of negotiation of meaning (Lynch & Maclean, 2001; Mackey, Kanganas, & Oliver, 2007; P. Robinson, 2011). Unlike negotiation of meaning, the opposite pattern was observed with negotiation of form. Interestingly, no negotiation on form was detected in the independent speaking task, making it contrast with the prediction of the Cognition Hypothesis which claims that increasing task complexity would lead to more interactional modifications, especially LREs (Gilabert, 2007; Gilabert, et al., 2009; Révész, 2011; P. Robinson, 2001, 2007, 2011). It remains unclear why the participants did not negotiate on form in the independent task, but one possible reason could be related to the participants’ attentional capacity and working memory. According to Skehan’s Limited Attentional Capacity Model (2001), humans have limited information processing capacity; therefore, in completing a task, learners are more likely to divert their attention to the content development of the task. Similarly, as suggested by VanPatten (1990), learners’ working memory is limited, thus they cannot equally pay attention to both meaning and form simultaneously. It is also important to note that in the 63 independent task, the participants were not provided with the initial input for their discussion. Therefore, the participants were required to brainstorm ideas on their own to complete the task. For the sake of task completion, it is possible that the participants prioritize meaning over form, which might possibly explain why they did not negotiate on form in the independent task. Moving to the quality interaction produced by the six interaction categories, it is interesting to find that negotiation of procedure was the most prevalent interactional feature, making up to one-third of the total interactional moves in the collected dataset. However, the quality interaction produced by this type of interactional feature remained the lowest, around 23-25%, which is far lower than the percentage of successful interaction produced by other interaction categories (ranging from 38 to 88%). This directly contrasts with Poupore’s (2004) findings in which negotiation of procedure was not the highest interaction feature but was the second highest in generating quality negotiations, up to 43%, nearly doubling the percentage of quality interaction produced by negotiation of meaning and form. The difference between this finding and that of Poupore’s (2004) study may be attributed to the differences in task types, learners’ proficiency levels and research contexts. While this study was conducted in an ESL context on two groups of different proficiency levels with the use of an integrated task and an independent task, Poupore’s (2004) study was conducted in an EFL context on a group of students ranging from intermediate to advanced levels with the use of different types of tasks (e.g., problem-solving prediction tasks, jigsaw tasks, and information-gap tasks). 64 According to Coughlan and Duff (1994), when learners are given a task, they themselves set their own strategies in terms of how they want to complete the task. In negotiating on task procedure, learners potentially produce pushed output. However, if they spend most of their time on the meta-talk of the task rather than the content of the task or the language involved in the task, the actual learning opportunities may be compromised, as indicated in this current study. Since the current study did not go further than collecting written responses from the participants, it remains unclear as to the reasons why the participants frequently negotiated on task procedure more than any other aspects. However, from the transcription of the participants’ interactional performance, the open-ended written responses, coupled with the close-ended analysis, particularly on the task direction aspect (which has the highest mean score compared to other examined aspects), it is hypothesized that the participants might have thought that they should keep them in line with the step-by-step task description for the sake of task completion. This is coupled with the fact that the participants were given clear instructions on the steps that they should take to complete the tasks, which might explain why they negotiated on task procedure significantly more often. Another interactional feature that is worth further discussion is negotiation of personal experience. The research findings of this study indicated that this negotiated feature only made up a small proportion of the total negotiation moves (10-12%). However, it was the second highest (only after negotiation of content) to promote quality interaction among the participants. This is consistent with what Poupore (2004) found in that during interactive tasks, learners often engage in conversation about their own lives 65 as a way of personalizing the task and making it more meaningful to them. Once learners begin to relate their personal experience to what is being discussed during task performance, they appear to be more engaged and more likely to experiment with the target language through output production. As engagement is one of the key factors to facilitate students’ second language acquisition (Larsen-Freeman, 2003), it is understandable that although the total amount of negotiation of personal experience was quite low (the second lowest), the percentage of quality negotiation of this type of interactional behavior was relatively high, with the successful ratio up to 50-60%. With regard to self-initiated repair, the research results indicated that there was not much difference in terms of the frequency of this interactional behavior across the two tasks, which contrasts the findings of Gilabert (2007) in which repair frequency increases with task complexity. If the Cognition Hypothesis on which Gilabert’s (2007) study based is true, it is expected that there would be more self-initiated repairs in the independent task. However, the current study found that the difference was insignificant. This would probably suggest that task types and task complexity may not affect the occurrence of this type of interactional feature, as pointed out in previous research that self-initiated repair may not be task-directed but rather be influenced by L2 proficiency, attentional control, and even L1 self-initiated repair behavior (Fox, Maschler, & Uhmann, 2010; Hellermann, 2009). This will be further discussed in the following section that examines the effects of learner proficiency on the occurrences of self-initiated repair. Overall, the integrated task and the independent task were common in terms of promoting interactional feedback and learning opportunities in several ways. First, it was 66 found that more negotiation of content and negotiation of meaning than negotiation of form irrespective of the task, as similarly found in other previous interaction studies (Abadikhah & Mosleh, 2011; Fuji, Mackey, & Ziegler, 2009; Kim, 2009; Kim & McDonough, 2008; Leeser, 2004; VanPatten, 2003; Watanabe & Swain, 2007; Williams, 1999, 2001). Second, negotiation of procedures appeared to produce the least quality interaction, in spite of being the most common type of interaction, which contradicts with what has been found in the literature (Poupore, 2004). More research is, therefore, needed to better understand the value of this interactional feature. Third, it is not clear whether the low rate of occurrence of negotiation of form in the given tasks was related to cultural background (as suggested by Philp, Walter, & Basturkmen, 2010), or other learners’ internal factors such as personality and gender (as suggested by Buckingham, 1997; D. Robinson, Gabriel, & Katchan, 1994), which requires more empirical research. Finally, it is found that self-initiated repair seemed not to be influenced by task types. Mediating Effects of Learner Proficiency on Task-based interactions Regarding how proficiency levels might mediate interaction-driven learning opportunities, the research results suggested that in general, the advanced learners appeared to produce significant more negotiation of content than their lower-level counterparts (12 as opposed to 6 times respectively), with the ratio of successful negotiation of the advanced group considerably higher than that of the lower group (83% and 50% respectively). Another noticeable pattern is that although learners of the two groups negotiated for meaning on the same regular basis (12 and 11 times), the percentage of quality interaction generated by the advanced group was twice as high as 67 that of the lower group (58% and 27%). This echoes with previous interaction research findings which found that as proficiency level increases, learners tend to negotiate more frequently and produce more modified output (Farrokhi et al., 2008; Kim & McDonough, 2008; Leeser, 2004; Mackey, 2013; Van Patten, 2003; Watanabe & Swain, 2007; Williams, 2001). This is because when learners become more proficient in a language, their attentional resources are freed up, which in turn enables them to notice more interactional feedback. On the other hand, the low-intermediate learners, with their limited cognitive resources, might have been constrained by their unstable developing linguistic system and produced less interactional feedback and fewer opportunities for modified output. Another possible reason for the obtained distribution of interactional features between the two groups of learners may be associated with learners’ working memory capacities (Kim, Payant, & Pearson, 2015; Mackey et al., 2010; Mackey & Sachs, 2012; P. Robinson, Mackey, Gass, & Schmidt, 2011). It is argued that in order to produce modified output in response to interlocutors’ feedback, learners have to go through several cognitive steps such as shifting their attention from meaning-focused to form-focused, identifying the error in the utterance preceding the feedback, and reformulating their initial non-target like utterance. All these steps require the ability to retrieve, reanalyze, and restructure stored information which are key components of working memory capacities. Therefore, it could be argued that working memory capacities, together with learners’ developmental levels, might have mediated the occurrences of interaction between the two groups of learners. 68 Despite producing less quality interaction than the advanced group, it is not to say that the lower level learners were not able to benefit through interaction. As indicated in Chapter 4, except negotiation of meaning and negotiation of procedure which led to relatively low quality interaction (around 25%), lower proficient learners were able to make use of interactional feedback and produce modified output, with the percentage of quality interaction up to 50% for both negotiation of content and negotiation of personal experience. This once again suggests that task-based interaction might be potentially beneficial to L2 learning, irrespective of the proficiency levels. However, the way that task-based interaction benefits L2 learning varies according to different contextual factors as well as learner-internal factors. Exceptionally, low-intermediate learners were able to successfully resolve all of the form-focused episodes (as illustrated in several examples in Chapter 4). It is worth noting that unlike previous research findings which found that advanced learners tend to pay more attention to formal aspects of the target language than lower proficiency learners (Kim & McDonough, 2008; Kim, 2009; Leeser, 2004; Van Patten, 2003; Watanabe & Swain, 2007; Williams, 2001), this research study found no significant difference in the occurrence of negotiation of form between the two groups of learners. It is possible that the small-scale of this study may not enable the researcher to detect the difference in negotiation of form between the two groups. Furthermore, it is interesting to note that the low-intermediate level learners produced twice as many negotiations of personal experience as advanced learners. As indicated in learners’ responses in the post-task questionnaires, during task performance, the lower-level learners had difficulty in finding ideas to complete the tasks. Therefore, it is speculated 69 that they were more likely to personalize the task by relating to their experience of the discussed topic as a strategy to keep the discussion going. However, whether this finding is an effect of learner characteristics (e.g., proficiency, or other underlying cognitive factors) or task characteristics (e.g., topic familiarity) warrants more empirical research. Another interesting finding from the dataset is that self-initiated repair was more common among advanced learners, with the number of self-initiated repairs produced by the higher proficiency group (15) nearly doubling that of the lower group (8). This finding is in sharp contrast with what has been reported previously in the literature that high self-repair frequency is an indicator of low proficiency, which is defined as grammatical and vocabulary encoding knowledge (Golonka, 2006; Hellermann, 2009; Segalowitz, 2010; Van Hest, 1996; Zuniga & Simard, 2018). It could be argued that the observed results might not be due to L2 proficiency but possibly to other underlying cognitive traits of learners such as working memory or attentional control which might have played a role in modulating the occurrence of self-initiated repair behavior. As suggested by Fincher (2006) and Simard, Bergeron, Liu, Nader, and Redmond (2016), learners who have higher working memory capacity are more likely to monitor and critically assess their own speech, which enables them to realize the gap in their language knowledge (e.g., non-target form in their own utterance) and make changes to close the gap. This might possibly explain why low-intermediate learners produced considerably fewer self-repairs than their more competent counterparts. Another plausible explanation could be due to the fact that the non-target forms do not impede meaning, therefore, there is no urgent need to repair them in real time conversation. It has also been pointed out 70 that L1 self-initiated repair behavior is also a significant predictor of L2 self-repair behavior (Derwing, Munro, Thomson, & Rossiter, 2009). However, this is beyond the scope of this current study. Relationships between Learners’ Perceptions of Tasks and Interaction-driven Learning With regard to the learners’ perceptions of the speaking tasks, strengths and weaknesses of the two speaking tasks were revealed. First, based on the learners’ comments in the post-task questionnaires, it is indicated that the task-based activities were fun and interesting, which supports the findings of Mackey, Fujii, BiesenbachLucas, Weger, Dolgova Jacobsen, Fogle, Lake, Sondermann, Kim, Tagarelli, Takada, Watanabe, & Abbuhl (2012) which found that learners generally had positive perceptions of task-based interaction. The descriptive statistics also showed that the study participants generally perceived the two speaking tasks to be beneficial for language learning. One major perceived benefit of the tasks was their facilitative role in enhancing learnerlearner interaction in classroom-based settings. This is reflected in the task design that enables the learners to share their ideas with their partners and negotiate with each other to reach an agreement on the issue posed in the task prompt. Another key benefit lies in the opportunities that allow the learners to notice the gap in their current linguistic knowledge, as supported by most of the learners’ remarks that while performing the communicative tasks, they realized that they could not express what they wanted to say. These findings support those of the previous interaction studies in that learners benefit from noticing through interaction (Schmidt & Frota, 1986; Swain, 1998). Although there 71 is still a debate about the effectiveness of noticing on L2 learning (see Truscott, 1998), it is generally believed that noticing the discrepancy between interlanguage and the target language is the first step to drive and direct learners’ attention to the linguistic aspects that they need to improve, which in turn may have a positive impact on their language learning process. According to Schmidt’s noticing hypothesis (1990), it is not until learners notice the language features in a meaningful context that they begin to reanalyze and restructure their interlanguage. Although no significant differences in learner perception were found between the two tasks, learners’ close-ended responses in the post-task questionnaires provided interesting insights. Specifically, the mean scores of task difficulty suggested that the independent speaking task was perceived to be more difficult than the integrated speaking task, although the difference was not statistically significant. In some ways, this corresponds to P. Robinson’s Cognition Hypothesis (2003) which predicts that tasks with no provision of ideas (e.g., an independent speaking task) are more challenging than those which provide learners with some initial ideas through input (e.g., an integrated speaking task). Another interesting finding is that, as suggested in previous research studies, the way learners interact and engage in the language learning processes is closely related to how they perceive the tasks they perform (Batstone, 2011; Dörnyei & Kormos, 2000; Mackey, 1999; Philp & Duchesne, 2008; Philp & Mackey, 2010; Schulz, 2001; Tarone & Bigelow, 2005; Watanabe & Swain, 2008). This study found that the learners perceived the integrated speaking task to be more interesting than the independent one. Having such a positive view towards the integrated speaking task may have pushed 72 learners to provide more interactional feedback and interactional modifications, which possibly led to more quantity and quality interaction in this task compared to the independent one, although the difference was insignificant. A closer analysis of learners’ open-ended responses also revealed that the majority of low-intermediate learners found it difficult to understand the content of the video input in the integrated speaking task. Having to process information in the listening input might have put a high cognitive load on learners, especially for those who are at a lower proficiency level. This points out to the fact that despite being perceived to be more interesting than the independent speaking task, the integrated speaking task did not automatically facilitate L2 learning, especially for the lower-level learners since they are required to process aural information to get access to the provided resource. However, it can be argued that once learners processed the provided input, their cognitive load became lighter as they were given initial ideas for the task discussion. This raises a pedagogical concern that despite being viewed as more interesting from the learners’ perspective, the integrated speaking task does not necessarily facilitate L2 learning equally for learners of different proficiency levels. With respect to task familiarity, learners’ close-ended responses showed that they were more familiar with the independent task than the integrated task. As predicted, the learners tended to produce more modified output when they are familiar with the tasks (Lynch & Maclean, 2001; Mackey, Kanganas, & Oliver, 2007; P. Robinson, 2011). The findings of this research support this claim by showing that there were more incidences of negotiation of meaning produced in the independent speaking task than the integrated 73 one. However, this current research did not lend support to previous research studies in terms of the provision of negotiation of form. As predicted, learners would pay more attention to the formal aspect of the target language in performing tasks with procedural familiarity (Mackey, Kanganas, & Oliver, 2007). However, no evidence of learners negotiating on form was found in the independent speaking task (which was considered more familiar than the integrated speaking task in terms of task procedure). While the conflicting findings highlight the need for more empirical research into the effects of task familiarity on the provision of interactional modifications across tasks, it is necessary to keep in mind that the very small number of learners in this research context made this study descriptive in nature and therefore the results should be interpreted with caution. All in all, the research findings were similar between tasks and groups with the exception of the followings. The independent task, as opposed to the integrated task, appeared to produce significantly more negotiation of meaning. Interestingly, negotiation of form was only found in the integrated task. For the two groups of different proficiency levels, the advanced learners tended to produce significantly more self-initiated repairs and negotiation on content compared to their low-intermediate counterparts. In relation to the literature, most of these findings were predicted. For instance, the participants barely negotiated on form, but rather negotiated on task procedure, task content and meaningrelated issues. However, the most interesting finding which contradicts the previous research studies is that despite being the most common interaction feature, the successful (quality) interaction of negotiation of procedure remained the lowest. 74 Chapter VI Conclusion Summary of Research Results This study investigated how task types and learner proficiency interact with interaction-driven L2 learning opportunities in an ESL context. With respect to the effects of task types on the occurrence of interaction-driven L2 learning opportunities, the research results indicated that the two interactive communicative tasks (i.e., an integrated speaking task and an independent speaking task) were equally beneficial in terms of promoting interactional features (e.g., negotiation of meaning, negotiation of form, negotiation of content, negotiation of procedure, negotiation of personal experience, and self-initiated repairs) which are known to facilitate L2 learning (Adams & Ross-Feldman, 2008; Pica et al., 1996; Poupore, 2004; Shehadeh, 1999; Van den Branden, 1997). A closer look at the distribution of interactional features revealed variation between the two tasks. While the independent task evidenced no negotiation of form, and more quantity and quality negotiation of meaning, the integrated task appeared to produce more negotiation of form. Apart from these marked differences, the two tasks showed no significant differences in the following ways. First, no significant difference was found in the frequency of self-initiated repairs, which raises a speculation that selfinitiated repairs might not be influenced by task types. Second, negotiation of procedure was the most prevalent interaction feature while negotiation of form remained the least common interaction move. Third, most of the quality interactions were produced out of negotiation of content while the opposite was true for negotiation of procedure. 75 In terms of how learners’ proficiency level may influence the occurrences of interactional feedback and interactional modifications, the research findings showed that the advanced learners appeared to produce significantly more negotiation of content than their lower-level counterparts (12 and 6 respectively). Also, self-initiated repairs appeared to happen more among advanced learners, with the frequency nearly doubling that produced by the lower-level learners (15 and 8). This is not consistent with what has been found in the previous research studies that self-repair frequency decreases as learners become more proficient (Golonka, 2006; Hellermann, 2009; Segalowitz, 2010; Van Hest, 1996; Zuniga & Simard, 2018). This raises an empirical question that high self-initiated repair frequency may not be primarily tied to proficiency but rather to other underlying cognitive traits of learners such as attentional control or even L1 self-initiated repair behavior as suggested in previous research studies (Derwing, Munro, Thomson, & Rossiter, 2009; Fincher, 2006; Simard, Bergeron, Liu, Nader, & Redmond, 2016). On the other hand, the lower level learners seemed to produce significantly more negotiation of personal experience compared to the advanced learners, with half of those leading to quality interaction. No significant differences were found in the rest of the interactional features. As for the learners’ perceptions of the use of the two interactive tasks, it is indicated that the learners as a whole generally had a positive attitude towards the tasks for promoting interactivity, engagement, noticing, and learner self-motivation. As commented by the learners, these tasks were not only interesting but also provided a rich learning environment, enabling them to share their ideas with their partners, negotiate to 76 reach a common goal, and help each other to reach a new understanding of the discussed topics. In addition, most of the learners, especially who were at the low-intermediate proficiency level, expressed their drive to learn more as they noticed the linguistic gap while trying to convey what they meant during their interaction. Although there is no guarantee that this drive will lead the learners to learn more, this is still promising on part of the learners’ perception. It can be argued that the information-exchange and convergent nature inherent in these tasks partly accounts for learners’ positive interaction-driven learning experience. However, several concerns were also raised regarding the challenges brought by the two tasks. While several learners expressed the difficulty in understanding the listening input in the integrated speaking task due to their limited vocabulary repertoire and the fast pace of the speakers in the videos, others found it difficult to think of their own ideas to complete the independent speaking task. Taken all into consideration, it can be argued that the two tasks were generally successful in providing the learners an authentic and engaging environment for L2 learning. Although it might be challenging for the low-intermediate level learners to complete the tasks (especially the integrated speaking tasks), this can be a positive sign as in order to make progress in the interlanguage development, learners need to be exposed to the input which is at a higher level than the learners’ current level of competence (Krashen, 1985). However, it is not to say that learners should be left to struggle with the tasks without any scaffolding from their teachers. This leads us to the following section where several pedagogical implications regarding how to make use of task-based interaction in L2 classroom settings are discussed. 77 Pedagogical Implications Based on the research findings, several pedagogical implications can be drawn in terms of the use of task-based interaction in a L2 classroom. The first implication of the research results is that, while there was evidence of learners negotiating on form in taskbased interactions, it was rather limited in the collected data. This is an unfortunate outcome as research studies have shown that attention to form is necessary for acquisition to take place (Schmidt 1990, 1994). The fact that learners rarely negotiated for form may be attributed to the task design itself as it did not give the students opportunities to focus on form. It is, therefore, suggested that task developers provide some space and direction for students to allocate their attention to the formal aspect of the target language. One possible suggestion is to incorporate a post-stage in the task cycle. For instance, after students discuss and reach the final agreement to address the task prompt, teachers can have them report their discussion findings to the rest of the class either in an oral or a written form. In knowing that they have to publicly present their discussion, learners would be more likely to pay their attention to form, whether it is grammatical, lexical or phonological aspects. Whether the post-task stage is done in a written or an oral form, it is important that students have a chance to analyze what they have done so that they can learn from experience and become better in other interaction practices. The second implication is that, as learners at the intermediate level pointed out that they had some difficulty in understanding the listening input in the integrated speaking task and finding ideas in the independent speaking task, it can be assumed that more scaffolding could help these intermediate learners to perform the tasks. One 78 possible implication is that teachers might pre-teach some challenging vocabulary related to the topic that learners are likely to encounter through a semantic mapping activity, which helps activate learners’ schema and helps prepare them for the upcoming task in the task cycle stage. A whole class discussion in which teachers and learners together brainstorm some ideas associated with the topic may also provide affordances for learner interaction. Another recommendation is that teachers train learners to be more proficient interlocutor to maximize the benefits of task-based interaction (Fujii, Mackey, & Ziegler, 2011). Through interaction-training sessions, learners are first introduced to the benefits of interactional processes on language learning. Then they might watch a video clip that demonstrates how simulated learners engage in a task-based interaction, followed by a whole-class analysis of interactional features that occur in the video. It is suggested that teachers pause the video whenever feedback is given so that learners might be made aware of different types of feedback (e.g., recast, negotiation of meaning, negotiation of form, etc.) as well as when and how to provide interactional feedback. Having understood how interaction helps promote L2 learning in a specific context, it is hoped that learners would become better interlocutors and provide more quality feedback when they get involved in task-based interaction. The third implication is that, although language teachers may have a desire for their students to speak fluently without preparation, it would be difficult for those who are nervous or need time to conceptualize what they are going to say and how they are going to say it to do this. The lower-level learners commented that they needed more time to take note of what they would like to say, which would, according to them, result in a 79 better conversation. This supports previous research that has highlighted the positive effects of planning time (N. C. Ellis, 2009; Ortega, 1995) and suggests that teachers give students enough time to think on their own before sharing their thoughts with their partners. This is not only helpful for reflective learners who need to think carefully about what they are going to say, but it also helps reduce learner anxiety. It is not to say, however, that preparation time should always be given. In preparing learners to deal with real-life communicative situations, learning activities which foster quick responses and fluency are sometimes necessary to be included in the lesson (Harmer, 2015). When students get familiar with such activities, they will hopefully gradually feel more confident in speaking spontaneously. Taken all together, while the integrated speaking task seems to be beneficial for learners in the sense that it familiarizes learners with authentic tasks they often encounter in real-life situations, the independent speaking task also provides a good learning condition where quality negotiation of meaning is fostered. Thus, in considering the potential benefits that the two task types bring to facilitate L2 learning, it is recommended that both of these two speaking task types be integrated into language learning curricula to give learners more opportunities for learning. Moreover, from a theoretical perspective, these speaking tasks are beneficial in the sense that they give learners an active role in their learning through meaningful interaction and collaborative learning. 80 Limitations of the Research and Recommendations for Future Research Despite the potentially new finding that negotiation of procedure might not contribute considerably to the provision of quality interaction, there are several limitations inherent in the study that must be acknowledged. The first limitation of this study lies in its relatively small scale. It is possible that the task effects on learner-learner interaction were not adequately detected because of the limited number of participants involved in this study (e.g., only two pairs at the advanced level and two pairs at the lowintermediate level), which made the research findings exploratory at best. Although the number of participants was limited, the research findings were in line with previous findings, which adds to the body of research showing that peer-peer interaction during interactive tasks brings a certain value to L2 learning. Moreover, by showing evidence that contrasts with the previous research in terms of the value of negotiation of procedure in generating quality interaction, it would be important to replicate this study with more participants across different proficiency levels as well as different educational contexts to find out if this finding holds true more generally. The second limitation of this study is that it only examined six verbal interactional features and did not take into account other non-verbal interactional features such as gestures, which have been recognized as an important part of human communication and may influence learner interaction and L2 learning outcomes (Faraco, Kida, McCafferty, & Stam, 2008; Goldin-Meadow, 2003; Gullberg, 2008; Kellerman, 1992; Sueyoshi & Hardison, 2005). Looking at the communicative uses of gestures might inform us about the process that is going on in learners’ mind during interaction which might be otherwise 81 missed by examining speech alone. Without examining non-verbal interactional features, the impact of task-based interaction on L2 learning could not be fully depicted. Future research is, therefore, recommended to expand the research scope by examining other interactional features, both verbal and non-verbal. Another limitation is related to the research instruments. Only two speaking tasks were used in this study. As indicated by previous research studies, different task types may engender different types of interactional feedback and modification. Future studies are in need to examine how other task types and task characteristics might influence negotiated interaction. Regarding the task design itself, since there was no post-test or post-task performance to examine if what learners attended to during task-based interactions was internalized, it is recommended to include follow-up tasks or tests in which learners have opportunities to report their discussion findings to the rest of the class to explore if the learning benefits gained from interaction remains in other contexts. Additionally, among a number of learner-internal factors, the current study only tapped into the mediating effect of learner proficiency levels on interaction-driven learning opportunities. More systematic research is warranted to investigate the impact of other learner variables (e.g., language anxiety, task motivation, gender, interlocutor relationship, etc.) on the occurrence of interactional features which are beneficial to L2 learning. Furthermore, it is believed that topics used for discussion might influence learner-learner interaction. It is, therefore, necessary to replicate the study with different topics ranging from different levels of familiarity to examine if the results remain the same or different from what obtained in this study. 82 References Abadikhah, S., & Mosleh, Z. (2011). EFL learners’ proficiency level and attention to linguistic features during collaborative output activities. The Iranian EFL Journal, 7, 179-198. Adams, R. (2007). Do second language learners bene t from interacting with each other? In A. Mackey (Ed.), Conversational interaction in second language acquisition (pp. 29–51). Oxford: Oxford University Press. Adams, R., Fujii, A., & Mackey, A. (2005). Research methodology: Qualitative research. Mind and context in adult second language acquisition: Methods, theory, and practice, 69-101. Adams, R., Nuevo, A. M., & Egi, T. (2011). Explicit and implicit feedback, modified output, and SLA: Does explicit and implicit feedback promote learning and learner–learner interactions?. The Modern Language Journal, 95, 42-63. Adams, R., & Ross-Feldman, L. (2008). Does writing influence learner attention to form? In D. Belcher & A. Hirvela (Eds.), The oral/literate connection: Perspectives on L2 speaking/writing connections (pp. 243–267). Ann Arbor, MI: University of Michigan Press. Batstone, R. (Ed.). (2011). Sociocognitive aspects of second language learning and teaching. Oxford: Oxford University Press. Bowles, M. A., & Adams, R. J. (2015). An interactionist approach to learner-learner interaction in second and foreign language classrooms. The Handbook of classroom discourse and interaction, 198-212. 83 Brown, H. D., & Lee, H. (2015). Teaching by principles: An interactive approach to language pedagogy (4th ed.). White Plains, NY: Pearson Education. Buckingham, A. (1997). Oral language testing: Do the age, status and gender of the interlocutor make a difference. Unpublished MA dissertation, University of Reading. Butler, F. A., Eignor, D., Jones, S., McNamara, T., & Suomi, B. K. (2000). TOEFL 2000 speaking framework. Princeton, NJ: Educational Testing Service. Bygate, M. (2009). TESOL and linguistics: Where’s the connection?. In J. Culpeper, F. Katamba, P. Kerswill, & R. Wodak (Eds.), English language and linguistics: Structure and use. New York, NY: Routledge. Bygate, M., Norris, J., & Van den Branden, K. (2012). Task‐Based Language Teaching. The Encyclopedia of Applied Linguistics, 1-8. Cao, Y., & Philp, J. (2006). Interactional context and willingness to communicate: A comparison of behavior in whole class, group and dyadic interaction. System, 34(4), 480-493. Chen, Y. S. (2016). The effect of learner–learner interactions on a multiple-choice discourse completion task. Learning, Culture and Social Interaction, 9, 55-67. Coughlan, P., & Duff, P. A. (1994). Same task, different activities: Analysis of a SLA task from an Activity Theory perspective. Vygotskian approaches to second language research, 173-193. 84 DeKeyser, R. M. (2011). Age effects in second language learning. In S. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition. New York, NY: Routledge. Derwing, T. M., Munro, M. M., Thomson, R. I., & Rossiter, M. J. (2009). The relationship between L1 fluency and L2 fluency development. Studies in Second Language Acquisition, 31, 533–557. Dörnyei, Z., & Kormos, J. (2000). The role of individual and social variables in oral task performance. Language teaching research, 4(3), 275-300. Dörnyei, Z., & Kormos, J. (2002). Motivational determinants of the quality and quantity of student performance in communicative language tasks. In Annual conference of the American Association of Applied Linguistics (AAAL). Doughty, C., & Pica, T. (1986). “Information gap” tasks: Do they facilitate second language acquisition?. TESOL quarterly, 20(2), 305-325. Duff, P. A. (1986). Another look at interlanguage talk: Taking task to task. In R. Day (Ed.): Talking to learn: Conversation in second language acquisition. Rowley, MA: Newbury House. Ellis, N. C. (2009). Optimizing the input: Frequency and sampling in usage-based and form-focused learning. In M. H. Long & C. Doughty (Eds.), The handbook of language teaching. Oxford: Blackwell. Ellis, R. (2001). Non-reciprocal tasks, comprehension and second language acquisition. In M. Bygate, P. Skehan, and M. Swain (Eds.): Researching pedagogic tasks: Second language learning, teaching, and testing. Harlow: Pearson. 85 Ellis, R. (2003). Task-based language learning and teaching. Oxford University Press. Ellis, R. (2009). Task‐based language teaching: Sorting out the misunderstandings. International Journal of Applied Linguistics, 19(3), 221-246. Ellis, R., & Barkhuizen, G. (2005). Analysing learner language. Oxford: Oxford University Press. Ellis, R., Tanaka, Y., & Yamazaki, A. (1994). Classroom interaction, comprehension, and the acquisition of L2 word meanings. Language learning, 44(3), 449-491. Faraco, M., Kida, T., McCafferty, S., & Stam, G. (2008). Gesture and the negotiation of meaning in a second language classroom. In S. G. McCafferty & G. Stam (Eds.), Gesture: Second language acquistion and classroom research, (pp. 280-297). London: Routledge. Farrokhi, F., Ansarin, A. A., & Mohammadnia, Z. (2008). Preemptive focus on form: Teachers' practices across proficiencies. Linguistics Journal, 3(2). Fincher, A. (2006). Functions of self-initiated self-repairs in an advanced Japanese language classroom. Unpublished dissertation, Griffith University, Australia Foster, P., & Skehan, P. (2009). The influence of planning and task type on second language performance. In K. Van den Branden, M. Bygate, & J. Norris (Eds.): Task-based language teaching. Philadelphia: John Benjamins. Fox, B. A., Maschler, Y., & Uhmann, S. (2010). A cross-linguistic study of self-repair: Evidence from English, German, and Hebrew. Journal of Pragmatics, 42(9), 2487-2505. 86 Fuji, A. & Mackey, A. (2009). Interactional feedback in learner-learner interactions in a task-based EFL classroom. International Review of Applied Linguistics in Language Teaching, 47, 267-301. Fujii, A., Mackey, A. J., & Ziegler, N. (2011). Metacognitive instruction and learning through task-based interaction. In 4th Biennial International Conference on Taskbased Language Teaching. University of Auckland, New Zealand. Gass, S., Mackey, A., Alvarez‐Torres, M. J., & Fernández‐García, M. (1999). The effects of task repetition on linguistic output. Language Learning, 49(4), 549-581. Gass, S., Mackey, A., & Ross‐Feldman, L. (2005). Task‐based interactions in classroom and laboratory settings. Language learning, 55(4), 575-611. Gass, S. M., & Mackey, A. (2007). Input, interaction, and output in second language acquisition. In B. VanPatten & J. Williams (Eds.), Theories in second language acquisition: An introduction, Mahwah, NJ: Lawrence Erlbaum Associates. Gass, S. M., & Varonis, E. M. (1994). Input, interaction, and second language production. Studies in second language acquisition, 16(3), 283-302. Gilabert, R. (2007). Effects of manipulating task complexity on self-repairs during L2 oral production. International Review of Applied Linguistics, 45(3), 215-240. Gilabert, R., Baron, J., & Llanes, A. (2009). Manipulating cognitive complexity across task types and its impact on learners' interaction during oral performance. IRALInternational Review of Applied Linguistics in Language Teaching, 47(3-4), 367395. 87 Goldin-Meadow, S. (2003). Hearing gesture: How our hands help us think. Cambridge, Harvard University Press. Golonka, E. (2006). Predictors revised: Linguistic knowledge and metalinguistic awareness in second language gain in Russian. Modern Language Journal, 90, 496-505 Gullberg, M. (2008). Gestures and second language acquisition. In P. Robinson & N. Ellis (Eds.), Handbook of cognitive linguistics and second language acquisition (pp. 276-305). New York, NY: Routledge. Harmer, J. (2015). The practice of English language teaching (5th ed.). Essex: Pearson Education. Hatch, E. (1978). Discourse analysis and second language acquisition. Second language acquisition: A book of readings, 2, 383-400. Hellermann, J. (2009). Looking for evidence of language learning in practices for repair: A case study of self-initiated self-repair by an adult learner of English. Scandinavian Journal of Educational Research, 53(2), 113-132. Iwashita, N. (1999). Tasks and learners' output in nonnative-nonnative interaction. In K. Kanno (ed.), Studies on the acquisition of Japanese as a second language. Amsterdam: John Benjamins. Julkunen, K. (2001). Situation-and task-specific motivation in foreign language learning. In Z. Dörnyei and R. Schmidt (Eds.) Motivation and second language acquisition. Honolulu, HI: University of Hawai’i Press. 88 Keck, C., Iberri-Shea, G., Tracy-Ventura, N., & Wa-Mbaleka, S. (2006). Investigating the empirical link between task-based interaction and acquisition. Synthesizing research on language learning and teaching, 13, 91. Kellerman, S. (1992). ‘I see what you mean’: The role of kinesic behaviour in listening and implications for foreign and second language learning. Applied linguistics, 13(3), 239-258. Kim, Y. (2009). The effects of task complexity on learner–learner interaction. System, 37(2), 254-268. Kim, Y. (2012). Task complexity, learning opportunities, and Korean EFL learners’ question development. Studies in Second Language Acquisition, 34(4), 627-658. Kim, Y., & McDonough, K. (2008). The effect of interlocutor proficiency on the collaborative dialogue between korean as a second language learners. Language Teaching Research, 12(2), 211-234. Kim, Y., & McDonough, K. (2011). Using pretask modelling to encourage collaborative learning opportunities. Language Teaching Research, 15(2), 183-199. Kim, Y., Payant, C., & Pearson, P. (2015). The intersection of task-based interaction, task complexity, and working memory: L2 question development through recasts in a laboratory setting. Studies in Second Language Acquisition, 37(3), 549. Kramsch, C. (1986). From language proficiency to interactional competence. The Modern Language Journal, 70(4), 366-372. Krashen, S. D. (1985). The input hypothesis: Issues and implications. Addison-Wesley Longman Ltd. 89 Lantolf, J. P. (2000). Second language learning as a mediated process. Language teaching, 33(2), 79-96. Lantolf, J. P. (2012). Sociocultural theory and the pedagogical imperative. In The Oxford Handbook of Applied Linguistics, (Ed.). Oxford University Press. Larsen-Freeman, D. (2003). Teaching language: From grammar to grammaring. Boston: Thomson/Heinle. Leaver, B. L., & Willis, J. R. (Eds.). (2004). Task-based instruction in foreign language education: Practices and programs. Washington DC: Georgetown University Press. Leeser, M. J. (2004). Learner proficiency and focus on form during collaborative dialogue. Language Teaching Research, 8(1), 55-81. Littlewood, W. (2007). Communicative and task-based language teaching in East Asian classrooms. Language teaching, 40(3), 243-249. Long, M. H. (1981). Input, interaction, and second‐language acquisition. Annals of the New York academy of sciences, 379(1), 259-278. Long, M. H. (1983). Linguistic and conversational adjustments to non-native speakers. Studies in Second Language Acquisition, 5(2), 177-193. Long, M. (1996). The role of the linguistic environment in second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), The handbook of second language acquisition. New York: Academic Press. Long, M. H. (2006). Problems in SLA. Second Language Acquisition Research Series. Mahwah, NJ: Lawrence Erlbaum Associates. 90 Loschky, L. (1994). Comprehensible input and second language acquisition: What is the relationship?. Studies in Second Language Acquisition, 16(3), 303-323. Lynch, T., & Maclean, J. (2001). A case of exercising: Effects of immediate task repetition on learners’ performance. In M. Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic tasks: Second language learning, teaching and testing, (pp. 141-162). Harlow: Pearson Education. Mackey, A. (1995). Stepping up the pace: input, interaction and second language acquisition. An empirical study of questions in ESL. Unpublished doctoral dissertation, University of Sydney. Mackey, A. (1999). Input, interaction, and second language development: An empirical study of question formation in ESL. Studies in second language acquisition, 21(4), 557-587. Mackey, A. (2002). Beyond production: Learners’ perceptions about interactional processes. International Journal of Educational Research, 37(3-4), 379-394. Mackey, A. (2013). Input, interaction and corrective feedback in L2 learning-Oxford Applied Linguistics. Oxford University Press. Mackey, A., Adams, R., Stafford, C., & Winke, P. (2010). Exploring the relationship between modified output and working memory capacity. Language Learning, 60(3), 501-533. 91 Mackey, A., Fujii, S., Biesenbach-Lucas, H., Weger, N., Dolgova Jacobsen, L., Fogle, J., Lake, K., Sondermann, K., Kim, K., Tagarelli, M., Takada, A., Watanabe, Y., & Abbuhl, R. (2012). Classrooms, FL and SL contexts, and task-based activities. In K. McDonough & A. Mackey (Eds.), New perspectives on classroom interaction in second language research. Amsterdam: John Benjamins. Mackey, A., Gass, S., & McDonough, K. (2000). How do learners perceive interactional feedback?. Studies in second language acquisition, 22(4), 471-497. Mackey, A., & Goo, J. (2007). Interaction research in SLA: A meta-analysis and research synthesis. In A. Mackey (Ed.), Conversational interaction and second language acquisition. New York: Oxford University Press. Mackey, A., Kanganas, A. P., & Oliver, R. (2007). Task familiarity and interactional feedback in child ESL classrooms. TESOL Quarterly, 41(2), 285-312. Mackey, A., Oliver, R., & Leeman, J. (2003). Interactional input and the incorporation of feedback: An exploration of NS–NNS and NNS–NNS adult and child dyads. Language Learning, 53(1), 35-66. Mackey, A., & Sachs, R. (2012). Older learners in SLA research: A first look at working memory, feedback, and L2 development. Language Learning, 62(3), 704-740. McDonough, K., & Mackey, A. (2008). Syntactic priming and ESL question development. Studies in Second Language Acquisition, 30(1), 31-47. McDonough, K., & Mackey, A. (Eds.). (2013). Second language interaction in diverse educational contexts (Vol. 34). John Benjamins Publishing. 92 Nuevo, A. M. (2006). Task complexity and interaction: L2 learning opportunities and development (Doctoral dissertation, Georgetown University). Nuevo, A. M., Adams, R., & Ross-Feldman, L. (2011). Task complexity, modified output, and L2 development in learner–learner interaction. In R. Robinson (Ed.), Second language task complexity: Researching the cognition hypothesis of language learning and performance, (pp. 175-202). Amsterdam: John Benjamins. Ockey, G. J., & Li, Z. (2015). New and not so new methods for assessing oral communication. Ortega, L. (1999). Planning and focus on form in L2 oral performance. Studies in second language acquisition, 21(1), 109-148. Ortega, L. (2005). Methodology, epistemology, and ethics in instructed SLA research: An introduction. The Modern Language Journal, 89(3), 317-327. Ortega, L. (2009). Understanding second language acquisition. London: Hodder Arnold. Ortega, L., & Iberri-Shea, G. (2005). Longitudinal research in second language acquisition: Recent trends and future directions. Annual review of applied linguistics, 25, 26-45. O’Sullivan, B. (2002). Learner acquaintanceship and oral proficiency test pair-task performance. Language Testing, 19(3), 277-295. Philp, J. (2003). Constraints on “noticing the gap”: Nonnative speakers' noticing of recasts in NS-NNS interaction. Studies in second language acquisition, 25(1), 99126. 93 Philp, J., & Duchesne, S. (2008). When the gate opens: The interaction between social and linguistic goals. In J. Philp, R. Oliver, & A. Mackey (Eds.), Child’s play: Second language acquisition and the younger learner (pp. 83-104). Amsterdam: John Benjamins. Philp, J., & Mackey, A. (2010). Interaction research: What can socially informed approaches offer to cognitivists (and vice versa)?. In R. Batstone (Ed.): Sociocognitive perspectives on language use and language learning. Oxford: Oxford University Press. Philp, J., Walter, S., & Basturkmen, H. (2010). Peer interaction in the foreign language classroom: what factors foster a focus on form?. Language Awareness, 19(4), 261-279. Pica, T. (1987). Interlanguage adjustments as an outcome of NS‐NNS negotiated interaction. Language Learning, 37(4), 563-593. Pica, T. (1992). The textual outcomes of native speaker-non-native speaker negotiation: What do they reveal about second language learning?. In C. Kramsch and S. McConnell-Ginet (Eds.), Text in Context: Cross-disciplinary perspective on language study, (pp. 198-237). Lexington, MA: D. C. Heath. Pica, T., & Doughty, C. (1985). The role of group work in classroom second language acquisition. Studies in second language acquisition, 7(2), 233-248. Pica, T., Lincoln‐Porter, F., Paninos, D., & Linnell, J. (1996). Language learners' interaction: How does it address the input, output, and feedback needs of L2 learners?. TESOL quarterly, 30(1), 59-84. 94 Pica, T., Young, R., & Doughty, C. (1987). The impact of interaction on comprehension. Tesol Quarterly, 21(4), 737-758. Plough, I., & Gass, S. M. (1993). Interlocutor and task familiarity: Effects on interactional structure. Multilingual Matters, 35-35. Porter, P. (1986). How learners talk to each other: Input and interaction in task-centered discussions. In R. R. Day (Ed.), Talking to learn: Conversation in second language acquisition, (pp. 200-222). Rowley, MA: Newbury House. Poteau, C. E. (2011). Effects of interlocutor familiarity on second language learning in group work. Temple University. Poupore, G. (2004). Quality interaction and types of negotiation in problem-solving and jigsaw tasks. In C. Edwards & J. Willis (Eds.), Teachers exploring tasks in English language teaching (pp. 242-255). Basingstoke: Palgrave Macmillan. Qiu, X., & Lo, Y. Y. (2017). Content familiarity, task repetition and Chinese EFL learners’ engagement in second language use. Language Teaching Research, 21(6), 681-698. Révész, A. (2011). Task complexity, focus on L2 constructions, and individual differences: A classroom‐based study. The Modern Language Journal, 95, 162181. Riccardi, D. (2014). Task Complexity, Task Type, and Learner-learner Interaction: A Replication Study with Adult ESL Learners (Doctoral dissertation). Robinson, D., Gabriel, N., & Katchan, O. (1994). Personality and second language learning. Personality and Individual Differences, 16(1), 143-157. 95 Robinson, P. (2001). Task complexity, task difficulty, and task production: Exploring interactions in a componential framework. Applied linguistics, 22(1), 27-57. Robinson, P. (2003). Attention and memory during SLA. In C. Doughty & M. Long (Eds.), The handbook of second language acquisition, (pp. 631-678). Oxford: Blackwell. Robinson, P. (2005). Cognitive complexity and task sequencing: Studies in a componential framework for second language task design. IRAL-International Review of Applied Linguistics in Language Teaching, 43(1), 1-32. Robinson, P. (2007). Task complexity, theory of mind, and intentional reasoning: Effects on L2 speech production, interaction, uptake and perceptions of task difficulty. IRAL-International Review of Applied Linguistics in Language Teaching, 45(3), 193-213. Robinson, P. (Ed.). (2011). Second language task complexity: Researching the cognition hypothesis of language learning and performance. Amsterdam: John Benjamins. Robinson, P., & Gilabert, R. (2007). Task complexity, the Cognition Hypothesis and second language learning and performance. IRAL-International Review of Applied Linguistics in Language Teaching, 45(3), 161-176. Robinson, P., Mackey, A., Gass, S. M., & Schmidt, R. (2011). Attention and awareness in second language acquisition. In S. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition, (pp. 247-267). New York, NY: Routledge. 96 Russell, J., & Spada, N. (2006). The effectiveness of corrective feedback for the acquisition of L2 grammar. In J. M. Norris & L. Ortega (Eds.), Synthesizing research on language learning and teaching, (pp. 133-164). Amsterdam: John Benjamins. Samuda, V., & Bygate, M. (2008). Tasks in second language learning. Basingstoke: Palgrave Macmillan. Schmidt, R. (1990). The role of consciousness in second language learning1. Applied linguistics, 11(2), 129-158. Schmidt, R. (1994). Implicit learning and the cognitive unconscious: Of artificial grammars and SLA. Implicit and explicit learning of languages, 22, 165-209. Schmidt, R. (2001). Attention. In P. Robinson (Ed.), Cognition and second language instruction. Cambridge: Cambridge University Press. Schmidt, R., & Frota, S. (1986). Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese. Talking to learn: Conversation in second language acquisition, 237-326. Schulz, R. A. (2001). Cultural differences in student and teacher perceptions concerning the role of grammar instruction and corrective feedback: USA‐Colombia. The Modern Language Journal, 85(2), 244-258. Segalowitz, N. (2010). Cognitive bases of second language fluency. New York: Routledge. Sheen, Y. (2008). Recasts, language anxiety, modified output, and L2 learning. Language Learning, 58(4), 835-874. 97 Shehadeh, A. (1999). Non-native speakers’ production of modified comprehensible output and second language learning. Language Learning, 49(4), 627-675. Shirvan, M. E., & Talebzadeh, N. (2017). English as a foreign language learners’ anxiety and interlocutors’ status and familiarity: An idiodynamic perspective. Polish Psychological Bulletin, 48(4), 489-503. Simard, D., Bergeron, A., Liu, Y. G., Nader, M., & Redmond, L. (2016). Production d’autoreformulations autoamorcées en langue seconde: rôle de l’attention et de la mémoire phonologique. Canadian Modern Language Review, 72(2), 183-210. Skehan, P. (2001). Tasks and language performance. Researching pedagogic tasks: Second language learning, teaching, and testing, 167-185. Skehan, P., & Foster, P. (2001). Cognition and tasks. In P. Robinson (ed.), Cognition and second language instruction, (pp. 183-205). Cambridge: Cambridge University Press. Spada, N., & Lightbown, P. M. (2009). Interaction research in second/foreign language classrooms. Multiple perspectives on interaction, 157-175. Storch, N. (2001). How collaborative is pair work? ESL tertiary students composing in pairs. Language Teaching Research, 5(1), 29-53. Storch, N., & Aldosari, A. (2013). Pairing learners in pair work activity. Language teaching research, 17(1), 31-48. Sueyoshi, A., & Hardison, D. M. (2005). The role of gestures and facial cues in second language listening comprehension. Language Learning, 55(4), 661-699. 98 Swain, M. (1998). Focus on form through conscious reflection. Focus on form in classroom second language acquisition, 64-81. Swain, M. (2005). The output hypothesis: Theory and research. In Handbook of research in second language teaching and learning (pp. 495-508). Routledge. Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied linguistics, 16(3), 371-391. Swain, M., & Lapkin, S. (1998). Interaction and second language learning: Two adolescent French immersion students working together. The Modern Language Journal, 82(3), 320-337. Swain, M., & Lapkin, S. (2000). Task-based second language learning: The uses of the first language. Language Teaching Research, 4(3), 251-274. Swain, M., & Lapkin, S. (2001). Focus on form through collaborative dialogue: Exploring task effects. In M. Bygate, P. Skehan, & M. Swain (Eds.): Researching pedagogic tasks: Second language learning, teaching and testing. London: Longmans. Swain, M., & Miccoli, L. (1994). Learning in a content-based, collaboratively structured course: The experience of an adult ESL learner. TESL Canada Journal, 12(1), 1528. Tarone, E., & Bigelow, M. (2005). Impact of literacy on oral language processing: Implications for second language acquisition research. Annual Review of Applied Linguistics, 25, 77-97. 99 Tarone, E. (2009). A sociolinguistic perspective on interaction in SLA. In A. Mackey & C. Polio (Eds.): Multiple perspectives on interaction: Second language research in honor of Susan M. Gass. New York, NY: Routledge. Tawfik, A. A., Giabbanelli, P. J., Hogan, M., Msilu, F., Gill, A., & York, C. S. (2018). Effects of success v failure cases on learner-learner interaction. Computers & Education, 118, 120-132. Truscott, J. (1998). Noticing in second language acquisition: A critical review. Second Language Research, 14(2), 103-135. Van den Branden, K. (1997). Effects of negotiation on language learners' output. Language Learning, 47(4), 589-636. Van den Branden, K. (Ed.). (2006). Task-based language education: From theory to practice. Ernst Klett Sprachen. Van den Branden, K. (2016). Task-based language teaching. The Routledge handbook of English language teaching, 238-251. Van Hest, E. (1996). Self-repair in L1 and L2 production. Tilburg: Tilburg University Press VanPatten, B. (1990). Attending to form and content in the input: An experiment in consciousness. Studies in second language acquisition, 12(3), 287-301. VanPatten, B. (1996). Input processing and grammar instruction in second language acquisition. Greenwood Publishing Group. VanPatten, B. (2003). From input to output: A teacher's guide to second language acquisition. McGraw-Hill. 100 Varonis, E. M., & Gass, S. M. (1985). Miscommunication in native/nonnative conversation. Language in Society, 14(3), 327-343. Vygotsky, L. (1978). Interaction between learning and development. Readings on the development of children, 23(3), 34-41. Wajnryb, R. (1990). Grammar dictation. Oxford: Oxford University Press. Watanabe, Y., & Swain, M. (2007). Effects of proficiency differences and patterns of pair interaction on second language learning: Collaborative dialogue between adult ESL learners. Language Teaching Research, 11(2), 121-142. Watanabe, Y., & Swain, M. (2008). Perception of learner proficiency: Its impact on the interaction between an ESL learner and her higher and lower proficiency partners. Language Awareness, 17(2), 115-130. Williams, J. (1999). Memory, attention, and inductive learning. Studies in Second Language Acquisition, 21(1), 1-48. Williams, J. (2001). Learner‐generated attention to form. Language Learning, 51, 303346. Willis, J. (1996). A flexible framework for task-based learning. Challenge and Change in Language Teaching, 52-62. Yuan, F., & Ellis, R. (2003). The effects of pre‐task planning and on‐line planning on fluency, complexity and accuracy in L2 monologic oral production. Applied linguistics, 24(1), 1-27. 101 Zuniga, M., & Simard, D. (2018). Factors influencing L2 self-repair behavior: The role of L2 proficiency, attentional control and L1 self-repair behavior. Journal of Psycholinguistic Research, 1-17. Ziegler, N., Ammons, S., Lake, J., Seals, C., Hamrick. P, Rebuschat, P. (2012). The developmental outcomes associated with interaction during conversation groups. In K. McDonough, & A. Mackey (Eds.): New perspectives on classroom interaction in second language research. Amsterdam: John Benjamins. 102 Appendices Appendix A Integrated Speaking Task Description Directions: You and your partner are going to watch a video about social media, but the content is different. 1. Watch your video two times and summarize the speaker’s point of view by writing down notes. 2. When both you and your partner are ready, discuss and share your summary of your speaker’s point of view. 3. Then together decide whether social media has more of a positive or a negative impact on people’s lives. Give at least 3 reasons for your decision. 103 Appendix B Independent Speaking Task Description Directions: You and your partner are going to discuss the topic: “Should students work while attending college?” 1. First, individually write down the pros and cons of this issue. 2. Second, share your list with your partner. Write down any of your partner’s ideas that were not on your list. 3. Third, together reach an agreement on whether you are in favor or whether you are against students’ having jobs while they are studying at college. Give at least three reasons for your decision. 104 Appendix C Questionnaire 1 This survey is conducted by a Teaching English as a Second Language (TESL) graduate student enrolled at Minnesota State University, Mankato, to better understand learners’ perception toward the integrated speaking task. This questionnaire consists of three sections. Please read each instruction and the example and then mark your answer. This is not a test so there are no “right” or “wrong” answers. The results of this questionnaire will be used for research purposes only so please give your answers sincerely. For those of you who wish not to participate, you can sit and do nothing or if this bothers you, you can spend a few minutes putting X’s on the page so it looks like you are participating. Thank you very much for your help! Section 1 In this part, I would like you to tell me how much you agree with the following statements by circling a number from 1 to 5 [Strongly disagree = 1, Disagree = 2, Neutral = 3, Agree = 4, Strongly agree = 5]. Please do not leave out any of the items. Example: If you like swimming very much, circle 5: Statement Strongly Disagree Disagree Neutral Agree I like swimming. 1 2 3 4 Statement Strongly Disagree Strongly Agree 5 5 Disagree Neutral Agree Strongly Agree 1. The task was interesting. 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 2. The task was challenging. 3. The directions for the task were clear. 4. The amount of time given was adequate. 5. This type of task should be used in academic speaking courses. 6. This task was similar to any speaking task you have done before. 105 7. This task helped you show your true English speaking ability. 1 2 3 4 1 2 3 4 5 8. This type of task should be used in placement tests for new international students at universities 5 in the US. Section 2 In this part, I would like to ask you for your opinion about the integrated speaking task. Please answer the following questions as much detail as you can and do not leave out any of the items. 1. What did you like about the task? 2. What did you dislike about the task? 3. How did the topic of the task affect your speaking? 4. What percentage of the listening input did you understand? a) 100 – 80% b) 79 – 60% c) 59 – 40% d) 39 – 20% e) 19 – 10% f) 9 – 0% 5. What were some problems you experienced during the task? 6. What suggestions do you have to improve the task? 106 Section 3 In this part, I would like to ask you for some information about yourself. Please provide the following information by writing your response in the provided space and putting a tick √ in one box that corresponds to your gender. 1. Name: ________________________________ 2. Major: ________________________________ 3. Age: ___________ 4. Gender: Female Male Other 5. Country of birth: _______________________ 6. First language: _________________________ 7. How many years have you been studying English? _______ years _______ months 8. How long have you been in the U.S.? _______ years _______ months 9. Your most recent English test scores: PBT TOEFL: ____________ CBT TOEFL: ____________ iBT TOEFL: ____________ IELTS: ___________ Other (Please specify:_____________): ___________ 10. Scores on speaking sections of English proficiency tests (if known or available): iBT TOEFL _____/30 IELTS _____/9.0 Other (Please specify:_____________): ___________ Thank you for your participation! 107 Appendix D Questionnaire 2 This survey is conducted by a Teaching English as a Second Language (TESL) graduate student enrolled at Minnesota State University, Mankato, to better understand learners’ perception toward the independent speaking task. This questionnaire consists of three sections. Please read each instruction and the example and then mark your answer. This is not a test so there are no “right” or “wrong” answers. The results of this questionnaire will be used for research purposes only so please give your answers sincerely. For those of you who wish not to participate, you can sit and do nothing or if this bothers you, you can spend a few minutes putting X’s on the page so it looks like you are participating. Thank you very much for your help! Section 1 In this part, I would like you to tell me how much you agree with the following statements by circling a number from 1 to 5 [Strongly disagree = 1, Disagree = 2, Neutral = 3, Agree = 4, Strongly agree = 5]. Please do not leave out any of the items. Example: If you like swimming very much, circle 5: Statement Strongly Disagree Disagree Neutral Agree I like swimming. 1 2 3 4 Statement Strongly Disagree Strongly Agree 5 5 Disagree Neutral Agree Strongly Agree 1. The task was interesting. 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 2. The task was challenging. 3. The directions for the task were clear. 4. The amount of time given was adequate. 5. This type of task should be used in academic speaking courses. 6. This task was similar to any speaking task you have done before. 108 7. This task helped you show your true English speaking ability. 1 2 3 4 1 2 3 4 5 8. This type of task should be used in placement tests for new international students at universities 5 in the US. Section 2 In this part, I would like to ask you for your opinion about the task. Please answer the following questions as much detail as you can and do not leave out any of the items. 1. What did you like about the task? 2. What did you dislike about the task? 3. How did the topic of the task affect your speaking? 4. What were some problems you experienced during the task? 5. What suggestions do you have to improve the task? 109 Appendix E Transcription conventions (Adapted from Poupore, 2004) 1. Split-second pausing is indicated by a coma (,). 2. Two or more approximate seconds of pausing are indicated by a series of dashes (-). For example, a two second pause is indicated by two dashes (--) and a five second pause is indicated by 5 dashes (-----). 3. Interruptions and/or overlapping speech are simply marked by a right-leaning slash (/). 4. Laughter is indicated by a star (*). 5. Unintelligible speech is indicated by a question mark within two brackets ([?]). One question mark approximately indicates one word of unintelligible speech, two question marks approximately indicates two words of unintelligible speech, and so on. 6. Sometimes, learners would mispronounce a word. When this occurred, the correct pronunciation spelling of the intended word meaning has been put inside two specialized brackets ({ }). 7. Interlocutor provision of backchannels are sometimes located within an interlocutor turn and are marked in between two parentheses ( ( ) ).