Running head: PEER TUTORING TO FOSTER SPOKEN FLUENCY IN ORAL TASKS Implementing Peer Tutoring to Foster Spoken Fluency in Computer-mediated Tasks

Twelve 14-to-15-year-old students participated in this investigaton. The participants faced difficulties to speak fluently in L2 and they evinced different levels of proficiency, as most of them ranked A1 or less and others A2. Therefore, a solution had to be found to provide students with an additional space to interact and develop their English language skills. Differences among learners were the starting point of this research and, consequently, peer tutoring was the strategy proposed. The benefits from collaborating in small groups are supported by previous studies. Tutors learned by teaching while tutees received more time for individualized instructions. Along with peer tutoring, oral tasks and computer-mediated instructions ornamented this intervention. Qualitative data from learning logs, video recordings, rubric for assessing spoken fluency and a final semi-structured interview pointed to the effectiveness of this inquiry. The results suggest that learners increased self-confidence, enhanced spoken fluency, improved the outcomes of tasks. Some of these results are also found in similar local and international investigations. What this paper adds is the use of computers as an additional tool for studentstudent interaction in distance, as well as the improvement of young learners’ spoken fluency in high school level.


Introduction Introduction to the study
The present study aimed at determining the possible influence of peer tutoring in students' spoken fluency when performing computer-mediated tasks. Nowadays, students face challenges in their learning of second language. One of the most common problems in the classroom is that students have different levels of proficiency in English. This situation arises, for instance, from governmental policies that allow a minimal percentage of students who are not promoted and, thus, pass without fulfilling the objectives of their course. Additionally, students who enroll in public schools are not usually classified according to their proficiency level of the target language. Having students with various levels of language proficiency in the same classroom creates a multi-level learning environment (e.g., multi-level classroom). Benson and Nunan (2005) claim that research has sought to answer why learners with similar cognitive capacities, under similar conditions, achieve different levels of proficiency; or why learners become diverse as they are engaged in second languages. The authors distinguish psychological factors such as motivation, affect and learning strategies; also, some social factors regarding the classroom and self-instruction. These differences in the classrooms, however, are not impediments for learners to benefit from their performances, for instance, in oral tasks.
Consequently, this research arises as an effort to yield strengths from weaknesses. Some years ago, Sharan and Shachar (1988) argued that the teacher-centered and traditional whole-class distribution of students in classrooms do not contribute to multi-level groups since, for example, learners with low proficiency might be unnoticed due to high achievers' continuous performance. Aspects such as self-motivation, self-regulation, selfmonitoring, initiative, positive self-image are commonly found in high achievers (Usuki, 2007).
In this regard, the present study sought to grant learners a central role through which they could work in small groups and help each other by alternating roles as listeners and speakers in order to have more time to ask for clarification or repetition, give instructions or receive individualized explanations that become negotiation of meaning that is considered a vital part of language acquisition (Council of Europe, 2001). Thus, this exchange of information might benefit high achievers and low achievers in their intention to master a second language since the former might reinforce their own learning by teaching, while the latter receive more time for individualized learning, repetition and correction.

Rationale of the study
The Colombian Ministry of Education (2006) grouped ninth grade students in B1, preintermediate level, according to the Common European Framework (CEF). In this stage, students should be able to take part in conversations about familiar topics such as leisure, dreams, school or work. Recently, the Ministry of Education also issued the Basic Learning Rights for secondary students (MEN, 2016) as a means to support the current curriculum development schools have.
This document agrees with the Colombian government's intention to foster language learning by means of the strategy Colombia Bilingüe. The Basic Learning Rights highlight that learning English is a progressive process where learners move from their current levels to more advanced ones and can be evinced, for instance, in the use of easier to more complex grammatical features (Harden, Witte & Köhler, 2006). In case of Colombian students, it is expected to reach preintermediate level (B1) over high school.
In spite of the aforementioned efforts, it is common to find that learners in the same group do not display the expected development when using the target language. Consequently, the educational community must seek to guarantee students their right to learn, in this case, a foreign language.

Needs analysis and problem statement
Twelve students agreed to participate in this study. In order to collect data to identify students' needs, the researcher used a survey, a proficiency test, and observation of students' performances in an oral task.
In the first part of the needs' analysis stage, students answered a computer-based survey to identify their motivations to learn English, activities they preferred to work on in the English class, tasks in which they have had high and low performance, and their preferences regarding individual or group work (see Appendix A). The results showed that students' major motivations to learn English were to have better job opportunities in the future, to travel abroad, and to enroll in a university. Additionally, conversations and role plays in pairs or small groups were the preferred activities to be worked in class by students. The participants also reported that memorizing vocabulary, speaking fluently and pronouncing correctly were some of the activities they found challenging in class.
Secondly, students took a proficiency questionnaire with fifty-five inquiries, from which thirty-five were based on the CEFR A1 level, and twenty questions based on the CEFR A2 (see Appendix B). In this part, the results showed differences in the levels of students' language acquisition. Four students ranked A2 with scores from 36 to 48 marks, while the others got 21 to 33 marks out of 54. In this sense, this group evinced different language proficiency levels although all learners belonged to ninth grade.
Thirdly, students worked in small groups and made videos of a conversation about family matters such as routines, families, schools, likes and dislikes. In this part, some students' salient behaviors were: students translated sentences into Spanish as a means to ask for clarification, hesitation was permanent in students' talk, students asked the teacher for vocabulary while they were performing the task, others preferred to use their first language or gestures to make themselves understood. In some cases, the expected outcome of the task was not achieved since some learners did not seem to count on strategies to overcome their lack of vocabulary or fluency during their performances.
The results in the needs' analysis showed a gap between students with high and low achievements. While some learners displayed aptitude toward learning English, others felt frustrated and experienced difficulties when performing oral tasks. In terms of proficiency level in L2, some learners in this investigation ranked A1 and A2, while others were below A1.

Justification of problem's significance
The Basic Learning Rights (MEN, 2016) state that ninth grade students must be able to "exchange information about academic and general interest topics in a conversation" (p. 21).
However, the participants of this study did not seem to be able to express their opinions or participate spontaneously in conversations since they did not count on enough strategies to overcome difficulties that arose during the interactions.
Additionally, according to the Colombian Ministry of Education (2006), students at B1 level are able to participate in conversations in which they express opinions and exchange information about personal topics or daily life. Nonetheless, due to school curriculum or teachers' methodologies, learners do not count on meaningful tasks that promote a balanced development of skills and focus mainly, for instance, on grammar achievement. In this regard, the Colombian Ministry of Education (1999) presents some curriculum development guidelines for language teaching in which language is seen as a tool to communicate and, for that reason, special emphasis should be made on its use in real settings, concentrating on fluency instead of specific grammar correction.
In terms of spoken fluency, the Council of Europe (2001) states that B1 learners can keep going comprehensibly and make themselves understood in short contributions by constructing phrases on familiar topics with sufficient ease to handle short exchanges, despite noticeable hesitation and false starts. Learners in this study displayed different proficiency levels of English that, in some situations, cause them to use mainly isolated utterances with much pausing what represents a gap between their expected and actual learning, as well as a breach between high achievers and low achievers (Hetzel & Soto-Hinman, 2009). It is paramount to find a proposal that involves receptive and productive skills improvement and bring students opportunities to succeed in learning a second language. For this reason, it was necessary to find out a possible solution to help students face challenges by working cooperatively.

Strategy selected to address the problem
To tackle the aforementioned challenges that students faced when interacting, peer tutoring was proposed since the literature and previous investigations pointed the effectiveness of this strategy. Vincent (1999) states that peer tutoring might extend the influence of instructions in classrooms thanks to high achievers' desire to help others; Harper and Maheady (2007) acknowledge increased opportunities for students to answer in small groups and receive individualized instructions; Tiwari (2014) says that active learning is fostered thanks to interactions in peer tutoring. In Colombian research, some of the findings regarding the use of peer tutoring are related to academic achievement and motivation (Ariza & Viafara, 2009;Cardozo-Ortiz, 2011;Nieto-Cruz, Cortés-Cárdenas, & Cárdenas-Beltrán, 2013).
Along with peer tutoring, tasks were planned in order to provide learners with meaningful opportunities to interact, negotiate and comprehend the meaning of language provided in the task input (Robinson, 2011a). Additionally, tasks invite learners to concentrate on spontaneous talk, that is, on meaning rather than on form (Willis & Willis, 2007). For the purpose of this study, tasks were found appropriate since they agree with Hinkel's (2017) conditions for fluency development: time devoted to be familiar with the material worked, focus on conveying meaning, time for practicing.
Since the arguments in favor of peer tutoring and tasks suggested by the literature might manipulate the reading of results in the present investigation, data collected was interpreted following the principle of grounded theory suggested by Corbin and Strauss (2015). Thus, any possible influence of the strategy selected to address students' difficulties emerged naturally.
As this inquiry intended to provide learners with additional opportunities to interact orally, the intervention of this project was held through online sessions so that students could take advantage of electronic devices. The use of computers agreed with students' inherent desire to incorporate these devices in classes, as it was manifested in the needs' analysis stage.
Peer tutoring, tasks and computers fitted students' needs since the participants with high and low achievement helped everyone by negotiating meaning when getting ready to perform the tasks. That is, the main task of the lesson was introduced by the researcher and, soon after, tutors and tutees got together to work in small groups. These notions are the bases of this research. The researcher intended to make the most of students' different levels of proficiency in English as well as their willingness to work in pairs or small groups to create opportunities to develop fluency through the suggested tasks.

Research question
In what ways does the participation in a peer tutoring strategy through computermediated tasks influence ninth grade students' spoken fluency at Colegio Reino de Holanda?

Research objectives
• To determine the possible influence that peer tutoring may have on students' spoken fluency when participating in computer-mediated tasks.
• To determine the increase (if any) in students' spoken fluency through the use of peer tutoring in computer-mediated tasks.

Conclusion
As it was stated before, peer tutoring could be a good solution to the difficulties in the commonly found multilevel class, taking into account factors such as learning backgrounds, levels of literacy, students' willingness to collaborate, among others. According to this strategy, diversity in classrooms arises as positive element for the learning practices by providing additional support and encouragement from peers rather than as a problem (Hess, 2001). By means of differences between tutors and tutees, students can overcome some difficulties and accomplish the goals of their learning as they count on meaningful opportunities to interact orally.
The following chapter presents the theoretical framework of the project, which includes relevant theories about peer tutoring and spoken fluency embedded in computer-mediated tasks.

Introduction
In this chapter, the three main constructs that underpin this research project are discussed: peer tutoring, spoken fluency and tasks. Theories and descriptions of several authors were brought to debate since their contributions are in support of earlier investigations and the present study. Additionally, descriptions of previous research related to the fields under investigation, at local and international levels, are considered in order to justify the research question and objectives.

Peer tutoring
Peer tutoring is an organized instructional method in which one learner tutors another in material in which the tutor is an "expert" and the tutee is a "novice" (Gordon, 2005). This definition, however, has been narrowed and adapted for the purpose of this study since, for instance, tutors in this investigation do not need to be experts in what they are tutoring, but they can be in the process of learning the material. In fact, Gillespie and Lerner (2008) argue that tutors do not need to be experts on the subject matter that is being worked on, or on grammar and correctness, but they do need to be experts in setting an appropriate environment for association and engaging with tutees. Similarly, Derrick (2015) draws attention to some characteristics tutors should have when coaching other learners: interest in helping people, openness to learning about tutees as well as the use of different techniques and ways to connect with them, and the ability to explain difficult concepts in simple, easy-to-understand language. Ryan, Reid and Epstein (2004) summarize three types of peer tutoring that are commonly present in classrooms. In the first, Classwide Peer Tutoring (CWPT), all students simultaneously participate in tutoring dyads. During each tutoring session, students can participate as both, tutor and tutee, or as only the tutor or tutee. The second type of tutoring, Cross-Age Tutoring, is useful for older students to extend their own knowledge by delivering instruction to younger learners who face learning problems in lower academic levels. Thirdly, Peer-Assisted Learning Strategy (PALS) is a version of CWPT where teachers identify learners with needs on specific skills and the appropriate pairs help them develop those skills. Dyads change regularly so that all students have the opportunity to be "coaches" and "players".
Several authors have highlighted the advantages of peer tutoring. Research indicates that peer tutors achieve significant academic and social gains, including improved self-esteem and self-efficacy, improved attitude toward school, and improved interpersonal functioning by instructing others, as well as confidence and opportunities to develop their own leadership skills (National Education Association, 2014;Phelps, 2007;Tiwari, 2014;Mastropieri, Scruggs, & Berkeley, 2010). Similarly, tutees receive individualized instruction, high level of academic achievement, confidence and comfort for receiving instruction from fellow students, and increased opportunities to respond in smaller groups (Harper & Maheady, 2007).
In spite of all the possible benefits of peer tutoring, this strategy is not commonly planned in classrooms. Some teachers might consider that implementing a peer tutoring program can be a very time-consuming practice for them, as well as an additional workload for students that already have the responsibility of their own learning. In same-age tutoring, some students may go off task as they are working with their friends. Moreover, if students only work with other students, peer tutoring would not promote a positive relationship between trainees and the teacher. Especially for L2 learners, in peer tutoring there is an intrinsic assumption that the peer tutor has more expertise in the L2 (Long and Doughty, 2009). This belief might lead some students to be reluctant to accept instruction from their peers. Others might argue that tutoring is a substitute for high quality instruction from teachers, or that they want to get what they pay for.
Nonetheless, in some situations, peer tutoring can occur on an incidental basis, especially on multigrade or multilevel classrooms, as a willingness of high achievers to help others, providing teachers with a powerful strategy for extending the instructional influence (Vincent, 1999). In this situation, the teacher is present for help, with the added benefit that there are tutors available for immediate company. Hence, peer tutoring should be conceived as collaboration among learners in order to achieve a common goal, usually the completion of a task or the answering of a question. This collaboration is evident, for example, with the "willingness to listen to others' ideas, suggestions, and opinions so that they can be discussed and integrated into further actions, such as decisions about how to complete a task" (Beatty, 2010, p. 109).
As peer tutoring might emerge spontaneously as a means of collaboration, research has shown that structured tutoring is more effective (Gordon, 2005;Jarvis, 2005;Topping, Buchs, Duran, & Keer, 2017). Garringer and McRae (2008) emphasize on four steps to build a practical peer tutoring program: considerations for the program, participant selection, training peer mentors, choosing match activities. In the first step, clear goals and a well-organized plan of action that impact on the identified students' needs are set. In this phase, observation, supervision, and feedback are critical to building a strong program. Secondly, the selection of participants, including mentors and mentees who are motivated and capable of providing help to their classmates, takes place. Thirdly, before the program begins and tutors instruct other learners, they should gain some training from teachers. In the last step, the authors recommend that the selected activities for the tutoring sessions promote friendship and trust, build developmental skills, and help tutees establish stronger bonds within the school environment.

Spoken fluency
One of the goals of learning an additional language is to use it naturally. In order to do so, Nation (2007) claims that four strands should be considered with an equal share of the class time, as well as learners' equivalent efforts: meaning-focused input, meaning-focused output, language-focused learning, and fluency development. Meaning-focused input refers to the development of receptive skills, listening and reading; while meaning-focused output is about how learners bring their thoughts to life by using speaking and writing skills. Regarding the third strand, learners must give deliberate attention to language features such as pronunciation, spelling, grammar and vocabulary. The fourth strand, fluency development, is maybe one of the most controversial in terms of what it involves since, sometimes, its definition is narrowed to speaking smoothly and rapidly without undue hesitations and pauses (Gatbonton & Segalowitz, 2005). Lennon (2000) defines fluency as the general ability of the spoken language demonstrated in "rapid, smooth, accurate, lucid, and efficient translation of thought or communicative intention into language under the temporal conditions of on-line processing" (p.26). In some proficiency testing, L2 speakers' oral achievement is judged against criteria such as pauses, silences and hesitations that, paradoxically, in the case of the mother tongue, are not seen as signs of disfluency but common tools to deal with turn taking, look for an appropriate word or organize the discourse. McCarthy (2010) considers an additional characteristic of fluency by suggesting that it implies the interactive production of two or more speakers who alternate speaker and listener roles and construct conjointly conversational discourse. Jones (2007) defines fluency as "being able to express yourself despite the gaps in your knowledge, despite the mistakes you're making, despite not knowing all the vocabulary you might need" (p. 18). This description of fluency agrees with the language proficiency of learners and the issue under investigation in this study since the participants are expected to be fluent with the resources they have within reach.
Defining fluency is a demanding task and, thus, measuring it is challenging too. Segalowitz (2010) distinguishes three notions of fluency, namely, utterance fluency, cognitive fluency, and perceived fluency. Utterance fluency can be measured objectively in a sample of speech having into account aspects such as words or syllables spoken per second or minute; use of short and long pauses and fillers such as uhs and uhms; repetitions, reformulations, and false starts (when speakers start to say something, stop, and start again using a different wording or idea); and the average number of words or syllables between pauses (Leonard, 2015). The second notion, cognitive fluency, cannot be measured directly like utterance fluency since it deals with the speaker's ability to efficiently plan and execute his speech, including content, vocabulary, and grammatical form. Regarding this second notion of fluency, Willis (2003) claims that "we make judgements of others on the basis of the language they use and the way they use it" (p. 19). That is, the perception of being a fluent speaker is based on other's ability to use a range of language forms and choose whatever seems appropriate to the circumstances.
Thirdly, perceived fluency is defined as the impression that listeners have of a speaker's fluency or a speech sample (Jong, Steinel, Florijn, Schoonen, & Hulstijn, 2013). Features such as speech rate and pauses, vocabulary, pronunciation, use of grammar, along with the perceived quality of interaction, are considered to quantify this listener's impression.
In this investigation, thus, fluency is measured using the parameters in the notion of perceived fluency and bearing into account the descriptors for A1 and A2 learners described by the Council of Europe (2001): All in all, fluency is an aspect of language that needs to be uncovered and fed through meaningful practice which may turn out in substantial improvements in learners' oral performance as speakers of L2. A fruitful strategy to foster such an attempt is the implementation of peer tutoring in computer-mediated tasks.

Tasks
The present study sought to identify the effects of peer tutoring in students' spoken fluency when taking part in computer-mediated tasks. An essential part of this inquiry is the use of pedagogical tasks that allow learners to interact in real-life settings using L2. Tasks are traditionally associated with everyday activities. So, in second language teaching, researchers usually use the term pedagogical tasks. Nunan (2004) states that a task is "a piece of classroom work that involves learners in comprehending, manipulating, producing or interacting in the target language" (p. 4). The author emphasizes that in tasks the intention is to convey meaning rather than to manipulate form. That is, special emphasis is made on fluency rather than on accuracy. Similarly, Robinson (2011a) claims that pedagogical tasks provide learners with opportunities to negotiate and comprehend the meaning of language provided in the task input or by the performance of a partner. The author highlights that tasks are appropriate to receive corrective feedback from teachers and peers. Willis and Willis (2007) identify two approaches in task-based teaching: form-based approach and meaning-based approach. In the first, special concentration on grammatical structures or functional realizations specified by the teacher is made. The success of this procedure lies on learners' capabilities to display these new specific forms with acceptable level of accuracy in their productions. In meaning-based approach, learners are encouraged to use the language as much as possible in spontaneous productions. Learners might, eventually, concentrate on language aspects in order to plan ideas and express themselves appropriately.
Later, teacher and students might focus on forms that occurred during the performance of the task just after its completion. In this approach, success is judged on whether or not learners communicate. This second approach converges with this research study because the participants are concerned with natural communication that, in turn, might develop spoken fluency. Ellis (2006) reviews three common stages that reflect the chronology of a task-based lesson. The first stage is 'pre-task' and is intended to establish the outcome of the task, to plan the time, or to prepare students to perform the tasks in ways that they boost acquisition, for instance, by completing a similar task or providing a model of how the task can be performed.
Secondly, in 'during task' stage, decision is made about whether students are allowed to use the input data while they perform the task, and to set strict time limits or allow students to complete the task in their own time. Interestingly, Yuan and Ellis found that students concentrated on accuracy when they were permitted to complete the task in their own time, but learners emphasized on fluency when time limit was set (as cited in Ellis, 2006, p. 85). Thirdly, the 'posttask' stage is to provide an opportunity for a repeat performance of the task with increased complexity, to encourage reflection on how the task was performed, or to focus on forms that seemed to be problematic to learners performing the task.
Additionally, for tasks to provide meaningful learning settings, Robinson (2011a) identifies some key features, namely, input, roles, settings, actions, monitoring, outcomes, feedback. Input refers to the written, visual or aural information presented for learners to work on the goal of the task. Roles are the functions that learners have in performing a task, information-giver and information-receiver. Settings are considered to prepare learners to communicate, including in and out of classroom grouping arrangements. Actions are the procedures to follow in performing and completing the tasks. Monitoring is the process to ensure that the task performance remains on track. Outcomes are the expected or evinced student's performance. Feedback includes the evaluation of overall learner's performance by the teachers or peers.
The researcher in this study was interested in using tasks since the evidence suggests they provide learners with time for purposeful interaction. Tasks help students improve oral communication skills, including improved linguistic and nonlinguistic aspects (Rohani, 2013).
Tasks benefit learners in relation to pronunciation, connected speech and use of fillers (Usma, 2015). Tasks support learners-initiated learning opportunities outside the classroom, thus, encouraging learner progress (Kangli, 2014). Tasks foster negotiation of meaning in web conferencing (Guo, & Möllering, 2016).

State of the art Previous research on peer tutoring
In the field of peer tutoring, several evidences have been collected regarding its positive effects on tutors and tutees' learning needs. Peer tutoring has been widely explored in different subjects including math, science and languages, at international and local levels. Many local studies, however, point at the effectiveness of peer tutoring in university students' general achievements such as good grades, motivation, and their influence in their decisions to stop dropping out of college (Ariza & Viafara, 2009;Cardozo-Ortiz, 2011;Nieto-Cruz, Cortés-Cárdenas, & Cárdenas-Beltrán, 2013). At the international level, peer tutoring has been broadly studied and several studies support their success with cross-age learners, students with disabilities, minorities, exceptional learners, among others. Doman and Bidal (2016) explored the experiences of peer tutors when instructing a group of undergraduate students in Macau, China. Data collected from focus groups and individual interviews demonstrated that peer tutors gained confidence and expertise in teaching when assuming a coaching role. On the other hand, peer tutors reported that building a relationship between peer tutors and tutees was imperative for the goals of peer teaching to be attained since this kind of tutoring might be demotivating due to trainees' early negative responses.
With the use of computers, Evans and Moore (2013) developed a web-based peertutoring system called Online Peer-Assisted Learning (OPAL). In this inquiry, students were asked to solve problems regarding topics related to class and, then, the student who showed competency by answering the problem correctly was given the opportunity to teach it. Students reported that different insights from different people helped them open their minds, and that they verified or deepened their own understanding by teaching others. This investigation brings new learning opportunities commonly found in face-to-face classrooms to online environments.
In Colombia, Viafara (2014) explored the effects of peer tutoring with a group of English as a foreign language university students. Fifteen advanced learners tutored their first semester peers for three semesters. The respondents provided data from surveys, journals and interviews.
Results suggested that tutors obtained positive outcomes from this practice. Firstly, the development of proficiency in English. Tutors reported that, by participating in the process of tutoring, their vocabulary was broadened and the provided feedback helped them improve their speaking in the foreign language. Secondly, tutors came to accept and value their limitations and were motivated to encounter opportunities to gain more preparation. These insightful gains are meaningful for learners when providing feedback to tutees in peer tutoring.
In relation to online peer work, Tolosa, Ordóñez, and Alfonso (2015) studied the effects of peer feedback with two groups of beginner eleven-year-old learners of English in Colombia and other learners of Spanish in New Zealand. Students were organized in pairs according to their proficiency in L2, and asked to provide feedback on peers' written texts. The study identified three categories of feedback in learners' computer-mediated responses to their peers: error correction, rewrite, and explanation.
The aforementioned results contribute to the present investigation considering that they display steps and relevant information in the pursuit of establishing a tutoring program. Although peer tutoring has been proven to be an effective strategy in different fields of education, as it was previously mentioned, Colombian education counts on little research about it. Hence, lots of investigation should broaden to schools and, especially, to second language acquisition settings for young learners.

Previous research on peer tutoring in computer-mediated tasks
Peer tutoring has been widely explored in the research field. Literature indicates that this strategy is effective in developing L2 learners' overall skills and its investigation has spread to computer-assisted learning settings. Davies (2000) researched computer-based peer assessment. Graham (2002) sought to create effective cooperative learning in face-to-face and virtual environments for distance education, focusing on creating the groups, structuring learning activities, and facilitating group interactions. McLuckie and Topping (2004) studied means of assessing the transferable skills needed for effective peer learning in online contexts.
Through constant revision of academic journals, Master's theses and online databases it was observed that, in Colombia, little research has been carried out to analyze the effects of peer tutoring on learners' spoken fluency in L2 and, specifically, the use of computers to congregate tutors and tutees still remains unexplored. Although, research related to peer work, spoken fluency or virtual learning environments might serve as scaffolding for the present study. Caicedo (2016) examined the effects of peer-correction and peer-assessment on 22 participants' spoken fluency. The implementation phase allowed learners to identify their own mistakes through the use of autonomous assessment and, then, provide feedback to their peers.
The participants reported that the strategies used in the study enhanced their perception of spoken fluency, increased self-confidence when speaking, and improved language awareness.
Similarly, Contreras (2013) analyzed the effects of the collaborative and self-directed speaking tasks on ten university students' spoken fluency. The results suggest that affective factors such as confidence and motivation are essential for students to speak fluently.
Additionally, working collaboratively makes students find ways to express their ideas more easily since peers help them shape language needed to accomplish communicative tasks.
Regarding the use of computers, in the local field, Castiblanco (2014) carried out a study using an online tool called Second Life. The study intended to explore the effects of task-fluency discussions in Second Life. The participants were 3 male and 5 female students aged between 17 and 19 years old. The results revealed that the selected strategy, along with the use of an appealing virtual environment, increased students' awareness of learning strategies, self-effort to speak fluently, and self-perception of being a fluent language user.

Justification of research question and objectives
In this investigation, learners evinced problems regarding their acquisition of L2, especially in their development of spoken fluency. In some situations, fluency is presented on the opposite side of accuracy since the former concentrates on meaning while the latter emphasizes on form (Council of Europe, 2001). Research suggests that spoken fluency progresses due to continuous exposure to language use and, therefore, 25% of class time and effort should be devoted to this characteristic of language (Nation, 2007). For that reason, peer tutoring in computer-mediated tasks emerges as a strategy that involves additional time for peer interaction based on real-life settings that emphasize on meaning rather than form.
However, after reviewing the literature and previous local and international studies to tackle similar problems, the researcher identified that, while the positive effects of peer tutoring in a face-to-face classroom are well documented, studies of computer-mediated peer tutoring are less common. Consequently, the present study sought to fill the existing gap regarding the use of peer tutoring with the aid of computers to foster students' spoken fluency and set the research horizon with the research question and objectives.

Conclusion
In multilevel settings, peer tutoring might be employed in many ways that benefit students with high and low achievements. Previous research, at local and internationsl levels, and authors who have contributed to the field of peer tutoring might serve as starting point to structure a plan. Peer tutoring can take place in contemporary online platforms that count on useful tools similar to those found in classrooms. Researchers might consider the key features of tasks to foster real interaction among learners as well as the preventive measures to use the Internet responsibly.

Introduction
The sections in this chapter include the type of study, the educational context, the participants' profile, the roles of the teacher-researcher, the data collection instruments, and the procedures of the implementation. The core ideas that shape the research design and its components are summarized below (Table 1). • While-stage: Learning logs, video recordings and checklist implementation.
• Post-stage: A semi-structured interview to a focus group was Conducted.

Type of study
The present study complies with the characteristics of Action Research (AR) since it allows teachers to inquire about problems with learning and reflect on their pedagogical practice.
Cohen, Manion, and Morrison (2007) define AR as "motivated by a quest to improve and understand the world by changing it and learning how to improve it from the effects of the changes made" (p. 298). Similarly, West (2011) argues that AR can help bridge the gap between theory and practice by addressing topics that are relevant to practicing teachers. Bearing this in mind, this study sought to analyze how the implementation of peer tutoring could improve students' spoken fluency when taking part in computer-mediated oral tasks.
AR is a flexible spiral process which allows action (change, improvement) and research (understanding, knowledge) to be achieved at the same time (Dick, 2002). Burns (2010)   In the planning stage, the researcher identified a problem and developed a plan of action to bring potential improvements. In this part of this research, a needs analysis was carried out and data was collected using a survey, a proficiency test and video recordings. This stage allowed the researcher to identify students' strengths and major troubles regarding their acquisition of L2, being spoken fluency the most noticeable one.
The second phase, action, was for the researcher to implement the plan through purposeful interventions into the teaching situation for a period of time. This intervention was organized thanks to the rapport provided by previous research and contributions that allowed the researcher to set the horizon of the investigation. The intervention took place from October 2016 to January 2017. Modifications made in this stage were related to helping tutors and tutees create bonds that allowed them to work confidently. suggestions about the instruments, especially from other researchers, were significant to think about the appropriateness of each tool and its connection to the research question.
In the reflection phase, the researcher evaluated and described the effects of the actions in order to understand the problem under investigation. This stage of the data analysis and further reflection took place in the first semester of 2017. Data was analyzed following the principles of grounded theory in order to avoid forcing information to yield envisioned results (Corbin & Strauss, 2015).

Context
This study took place at Colegio Reino de Holanda, a public school in Bogota, the capital city of Colombia. This school is divided into three campuses and counts on elementary, middle and secondary education. Recently, the school adopted the full-day shift for secondary students and allocated four academic hours a week to English class. One major issue in this school is that its members have to constantly deal with social problems like aggression and intolerance.
Regarding the syllabus, at the end of the academic year, English teachers set out adjustments according to their class experiences, students' difficulties, Colombian policies and guidelines for language teaching. Unfortunately, most of the goals for ninth graders mainly address grammar achievements and, thus, students do not count on meaningful communicative settings.
With regards to the sessions, the present study was developed with the aid of an online classroom platform called Blackboard Collaborate. This application was published by Blackboard Inc. and is intended to provide web-conferencing for distance teaching and learning.
This software is appealing since it is user-friendly and counts on characteristics found in common classrooms that allow users to share documents and presentations, chat, arrange students in small groups, and record the sessions on video, among many other features.

Participants
Twelve ninth graders agreed to participate in the needs analysis stage of this study. In this part, most of the students expressed that learning English was paramount for enrolling in a university or getting better job opportunities in the future; however, their contact with English was limited to school. The selection of students' roles in the investigation, tutors or tutees, was based on their performances and proficiency level in English manifested during the needs analysis sessions and their genuine interest in becoming tutors. The initial findings regarding students' interests and challenges are better described in chapter 1.
Regarding the socioeconomic status, the students in the institution belong to the socioeconomic strata 1 and 2. They live in Marco Fidel Suarez neighborhood in the south of the city and its surroundings. Their parents mainly work in informal jobs or own independent small shops.
In terms of their linguistic needs, students' more significant challenges are related to oral interaction. Although students have L2 background and are able to express basic ideas in the target language, they fail when joining discussions, producing spontaneous talk, finding the appropriate vocabulary to speak their minds, or designing a plan for a speech.
In relation to materials for the English class, students are not asked to buy any books since it was established in the fee-free policy for public schools. Teachers create or adapt materials related to the topics in the syllabi and bring them to class. Additionally, students have access to a bilingualism room, electronic boards, computers and tablets used in the classrooms.

Researcher's role
Teachers conducting action research are not only observers or explorers of the situation but become active participants in the investigation (Burns, 2010). In this study, the researcher adopted various roles such as facilitator, observer and participant. The teacher-researcher trained students in using peer tutoring strategy to overcome the difficulties they faced in oral interactions. Besides, the researcher was acquainted with language learning theory, which guided his work toward effective solutions to the issues that concern this inquiry. The roles of active participant and observer were evident while the action plan was implemented and the data gathered in order to answer the research question.

Ethical considerations
In this study, the researcher considered that the participants were underage. For this reason, consent forms were signed by the school's principal and parents authorizing the researcher to conduct the project (see Appendix C). The participants were informed about the right to voluntarily accept or refuse to be in the process and that the information gathered would not have any negative influence on their grades.
Additionally, students and parents signed an Acceptable Use Policy (AUP) document with a detailed description of students expected behaviors when using computers, tablets or mobile phones (see Appendix D). This AUP intended to make students aware of the basic Internet safety rules before they surfed independently.
Regarding the data collection stage and the analysis of information in this research project, the participants' anonymity and confidentiality was maintained at all stages.
Additionally, the researcher bore in mind Creswell's (2012) ethical considerations to honestly report the findings that emerged from the analysis of data avoiding biased information or influenced by personal interest. The participants in this inquiry were provided with information about the findings in this investigation for them to read and understand. Finally, credit for authorships was given when necessary.

Data collection instruments
In order to collect data, this study used learning logs, semi-structured interview, video recordings, and a checklist for assessing spoken fluency.

Learning logs
The first instrument to gather data in this research refers to written entries in learning logs. This instrument is widely accepted in qualitative investigations since they represent a description of actual situations that occur in learning environments and are described by students, the main participants of this intervention. According to Hurd and Lewis (2008), one key element of learning logs is the appropriateness to keep record of one's own activity in order to learn from that experience and shape the upcoming phase of learning.
Another possible benefit of learning logs is mentioned by Cunningham, Dawes and Bennett (2004). The authors highlight that logs force learners to be explicit about what their learning consists of, and that being aware of that education process might lead to intentional learning. In this regard, learning logs were used in the first four sessions of the implementation stage, hence, learners were asked to record their experiences, feelings and reactions in learning logs as a means to keep track of their performances, shape their behaviors or concentrate on specific tools that they might find useful in future tasks (see Appendix E).

Recordings
Video recordings were used as the second instrument to collect information in this investigation. As a pedagogical tool, these recordings can help learners to assess and improve their speaking abilities because students are able to identify those aspects in their speech that are hindering effective communication or those which simply need reinforcement (Ariza, 2003). Koshy (2005) highlights some advantages of using video recordings in research such as the capture of students' behaviors and attitudes with greater accuracy, records can be viewed and reviewed, it is possible to carry out studies which need a sustained period of development and data collection so as to note changes.
In this study, four video recordings were accomplished by learners in order to provide themselves and the researcher with an overall view of their fluency development before, during and after the implementation. That is, one video was firstly recorded in the planning stage as a means to characterize students. Later, during the implementation stage, learners were recorded once every three sessions while performing the suggested task in the class. In the observation stage suggested by Burns (2010) for action research, one video was recorded in order to have an instrument that directly accounted for learners' spoken fluency progress, these video recordings were analyzed by peer tutors and the researcher with a checklist.

Checklist
An objective and structured method of gathering data involves using a checklist. A checklist can be useful for the researcher to focus on a limited amount of particular actions that do not change while the observations are being made. In fact, Jackson (2008) defines checklist as "a tally sheet on which the researcher records attributes of the participants and whether particular behaviors were observed" (p. 85).
In this investigation, a checklist was considered as an important tool in the standardization of the assessment of students' spoken fluency when they were performing oral tasks (see Appendix F). Peer tutors and the researcher evaluated tutees' presentations with a list of criteria that could be marked as present or absent. As the checklist was used four times in the data collection stage, in every phase of action research, it helped to keep track of every student's progress; therefore, the improvement could be determined.

Interview
The fourth instrument to gather data was a thirty-minute semi-structured focus group interview conducted in Spanish to allow the participants to answer the questions freely and spontaneously (see Appendix G). Wengraf (2001) says that, under the right conditions, semistructured interviews can yield richer data than fully structured ones since the interviewer uses predetermined questions, but also allows for expansion and elaboration in the responses by asking additional questions to follow up on any interesting answer to the standard questions.
One of the advantages of focus group interviews is that, since they involve several informants, interviewees can stimulate and be stimulated by each other and, thus, the researcher might elicit richer data than when conducting individual interviews (Nunan & Bailey, 2009).
Interviews can give students opportunities to reflect on their feelings and behaviors that, for instance, researchers identified during a specific event of investigation. In this regard, for this research project, the researcher highlighted some students' salient behaviors in the videos learners recorded during the tasks, replayed their presentations, and addressed some questions to see whether students considered the use of peer tutoring as a means to overcome the challenges they faced when taking part in the tasks. This instrument allowed the researcher to collect data after some cycles of AR had been completed and served to gather specific information that related to categories tha were emerging.

Validation and piloting
In collecting data for AR, triangulation arises with an important role. Triangulation is defined by Cohen et al. (2007) as "the use of two or more methods of data collection in the study of some aspect of human behavior" (p. 141). Then, data can be compared, contrasted and crosschecked to see whether what the researcher found through one source is backed up by other evidence. In the case of this study, it refers to the responsibility the researcher had to find ways to strengthen the data collected in order to give a possible answer to the research question avoiding biased assumptions. Nunan and Bailey (2009) state that triangulation is a quality control strategy that allows the researcher to gain solid insights by taking into account more than one source of data. Thus, triangulation of the instruments ensured the validity and reliability of this investigation. In order to validate this research design, the researcher administered these instruments for data collection: learning logs, video recordings, semi-structured interview, and a checklist for peer tutor to assess tutees' performances. The instruments were piloted with ninth-grade learners in order to find possible refinements and changes to wording before gathering data. Furthermore, the researcher reviewed the appropriateness of each instrument and their connection with the research question in order to collect relevant data for the support of the research study.

Conclusion
The research design of this study allowed the researcher to structure a path for learning about the effects of peer tutoring in learners' spoken fluency. In this chapter, the participants were described as well as their roles in this inquiry. Learning logs, video recordings, a checklist, and a semi-structured interview in focus group were chosen to gather data and their connection to the research question and objectives were explained. The next chapter will present the pedagogical implementation and intervention of this study.

Introduction
The main focus of this research involved the implementation of peer tutoring to explore its effects in students' speaking fluency when taking part in computer-mediated tasks. As it was explained in chapter one, the participants in this investigation showed different levels of proficiency in L2. With this in mind, peer tutoring was the strategy proposed as a possible vehicle to level students' performance in, especially, oral tasks since during the needs analysis stage, the participants confirmed that they felt interested in taking part in oral activities through group work. However, for peer tutoring to produce the expected outcomes, a detailed picture of the pedagogical procedures framing the study was required.
This chapter presents the visions of language, learning and curriculum as they are seen for the purpose of this research. Moreover, it describes in detail the process and steps followed in the pre, during and post implementation stages. The data collection instruments used in each stage, the objectives of the tasks in the implementation and the materials used are specified.

Vision of language
The role of language in human life is definitely unique. However, we might try to compare certain characteristics of it to different things such as tools, feelings, music, etc. In this research, language is seen as a key that has allowed the human race, throughout the years, to overcome barriers and access new opportunities. Nowadays, for instance, most of the candidates that apply for a position, different jobs or scholarships must evidence knowledge of a second language.
This conception agrees with the needs that the participants in this investigation have. One of the major reasons for them to learn English is because it opens job and academic opportunities. With this in mind, language raises students' awareness of their own attitudes, values and beliefs as a means to accept different cultures and languages, and to overcome any barriers about superiority or inferiority (Jones, 2013;Perez, 2003).

Vision of learning
In this investigation, learning is perceived as a collaborative effort. The learning process utilizes students' uniqueness and their abilities to work collaboratively in order to fulfill tasks.
Sometimes an outstanding group is a talented group of individuals. In this sense, each student is firstly in charge of his own learning process and his capabilities to face threatening situations, which evidences autonomy.
When individual skills are identified, group work could be displayed successfully.
According to Breen and Mann (1997), group work is the natural medium for negotiation and is part of an autonomous learner. McMahan (1998) mentions that teams need to have considerable interaction and must work together to complete assigned projects. This collaboration requires effective communication and information sharing among the members of the team. Thus, this research project intended to use peer tutoring as a medium for students to contribute when they are asked to work in groups, and learn how to articulate some of their personality characteristics as well.

Vision of curriculum
This research project sees real life situations as the starting point of lessons. Benson (2006) states that teachers should take into account students' contexts and out-of-class activities in which they tend to be engaged. It could be difficult for learners to appreciate something they do not have in their surroundings. If students feel identified with what they are supposed to learn, they will surely show positive outcomes.
Another element that is taken into account in this investigation is students' inherent interest in technology. In language learning, teachers should consider the wide range of options that technology can bring into our classes. Unfortunately, more emphasis has been put on the purchasing of electronic devices than on teaching students how to take advantage of them, and avoid being exposed to potential risks. Kizza (2003)

Instructional Design
The pedagogical implementation of this research took place for a total of ten sessions, each one consisting of a two-hour class. The sessions were framed within the task-based language learning methodology. In this regard, each lesson included a pre-task, during task, and post-task phases, as suggested by Ellis (2006). The lessons were set to demonstrate the influence of peer tutoring on the speaking fluency of a group of ninth graders.

Description and rationale
Taking into consideration the results obtained in the needs analysis stage of this investigation, the respondents were classified into three levels (above, at, and below) stated by Roberts (2007) as follows: 4 students in the above level, 2 students in the at level, and 6 students in the below level. The planned goal of fostering speaking fluency through peer tutoring strategy led the researcher to organize students in small groups that included one peer tutor and two tutees.
Regarding the sessions, the activities were planned following the task-based teaching methodology. Ellis (2006) advocates three phases that are given for a task to be executed: pretask, during task, post-task. These moments were mirrored in the development of each session. In the 'pre-task' phase, the topics were introduced through selected video, readings or pictures exposure and students were explained the purpose and utility of the tasks, or worked on non-task preparation activities such as brainstorming, mind-maps, or guided questions in order to activate prior knowledge. In the 'during task' phase, students were given time to get in small groups, one peer tutor and two tutees, and were allowed to use the previous materials to be ready to perform the tasks. In this phase, students' performances were recorded on videos once every three sessions. In the 'post-task' phase, students were invited to reflect on how the task was performed by using learning logs (see Appendix E). Also, this time was devoted to reviewing forms that seemed problematic to learners when performing the tasks. Additionally, peer tutors watched the videos that tutees had previously recorded and evaluated their work using a checklist (see Appendix F).
The content of this study was planned keeping in mind the school syllabus for ninth graders. The selection of attractive topics that agreed with students' interests was the starting point of the sessions, as well as tasks where participants acted primarily as language users in the sense that they were able to use the same kind of communicative processes as those involved in real-world activities (Ellis, 2003). Each lesson was prepared bearing in mind students' capabilities toward completing the tasks as they were progressively more challenging. Videos, pictures, surveys, puzzles and readings were used in order to support the implementation of this study (see Appendix H).

Description and rationale
This implementation was developed in three stages which were: pre-, during, and postimplementation (Table 2). Analyzing if peer tutoring strategy was helpful during the accomplishment of the tasks.
Students video recordings of their performances in oral tasks.
Peer tutors' evaluation of tutees' work.

Post-One week
Collection of students' insights regarding the effects of peer tutoring on their spoken fluency.

• Interview
In the pre-implementation stage, the researcher sought to identify students' needs and language interests through the use of a survey and a proficiency questionnaire. Learners also performed and video-recorded an oral task about a familiar topic that revealed some difficulties when interacting orally, for instance, some learners did not seem to count on strategies to overcome their lack of vocabulary or fluency during their performances. Students showed different levels of proficiency in the acquisition of L2 which influenced the researcher to set the general horizon of the research study. In this phase of the investigation, the researcher sent consent letters to learners' parents and to the school principal in order to contextualize the subject under investigation and describe how the problem was going to be tackled (see Appendix C).
In the while-implementation stage, the researcher trained high achievers to assume the role of peer tutors during the online sessions devoting two forty-five-minute sessions to teaching and modeling as suggested by Spencer (2006). The researcher also observed students' performances during the tasks and the appropriateness of the selected strategy to address the problem. Every lesson at this stage included three parts: pre-task, during task and post-task. In the sessions, students were mainly presented the goals of the class and useful vocabulary and, then, peer tutors, and tutees got in small groups (three to four members), each team consisting of at least one pupil who is high performing, one who is average performing, and one who is low performing, in order to be ready to complete the suggested tasks. Additionally, the researcher gathered data using learning logs after every session for students to reflect upon their performance during the tasks, as well as a checklist for peer tutors to assess their tutees achievements during the recorded tasks.
The post-implementation stage served for students to take part in a semi-structured interview with the researcher regarding their perceptions of the effects of peer tutoring on their speaking fluency. Furthermore, this phase was devoted to the analysis, validation and the interpretation of the data collected, and to sharing findings connected to the research question and the objectives.

Conclusion
This chapter described the processes followed to conduct this research study, including the visions of language, learning and curriculum. Here, the researcher described how peer tutoring and computer-mediated tasks converged thanks to the theoretical principles delineated in chapter two. Additionally, the instructional design for the stages and the instruments used to gather data during the process of implementation were explained according to the researcher's perspective. However, modifications are welcome according to the learners' needs and preferences. It is advisable that researchers become aware of the possible problems they might face in each lesson and be ready to modify, adapt, or look for appropriate material for the best development of the phases.

Introduction
This chapter describes the analysis of the influence of peer tutoring on ninth grade learners' spoken fluency through the use of computer-mediated tasks. The data analysis methods and procedures carried out in the research project, including the findings emerging from data, are presented in order to answer the research question.
In this section of the paper, the information gathered from the implementation stage is consolidated and analyzed using the principle of grounded theory suggested by Corbin and Strauss (2015) for data analysis in qualitative research. According to this methodology, the phenomenon under investigation needs to be inductively interpreted through salient themes or categories that are grounded in data, in order to avoid imposing theories coming from literature, previous experiences, or anticipated results. Thus, the researcher's ability and sensitivity to present the participants' perspectives through the immersion in data was paramount.

Data management procedures
In this study, four instruments were used to gather data: learning logs, video recordings, a rubric to assess the oral production of learners, and a semi-structured interview.
Learning logs were used at the end of four sessions. This instrument was implemented to record learners' experiences, feelings and reactions with regard to their roles as tutors and tutees when working in group tasks. The participants were provided with a form with guiding questions for them to fill out. Learning logs yielded enriching data directly from students' records of their performances, feelings and goals for the upcoming classes.
Secondly, students' performances were recorded on video once every three sessions in order to gather evidence regarding the changes in their spoken fluency. Videos were later revised by the researcher and tutors using a rubric that focused on the perceived fluency of learners.
Vocabulary and expressions, use of grammar structures, confidence, learners' ability to focus on tasks, intonation and pronunciation were the criteria considered to evaluate the participants' development of spoken fluency. This rubric was used at the end of the implementation process to appraise the possible effects of the strategy proposed.
Finally, a semi-structured interview allowed the researcher to elicit final insights regarding students' perceptions toward the implementation of peer-tutoring in computermediated tasks. Tentative questions were prepared beforehand and others were added along the discussion.

Data analysis methodology
The present qualitative research involved a specific methodological approach for the analysis of the data called grounded theory. Grounded theory is defined as an interactive process with data in order to build emerging theory that fits the reality that it represents (Corbin & Strauss, 2015). This interaction enables the researcher to classify the relevance of the information and generate theory based on reasoning (Cohen, Manion & Morrison, 2007).
In using grounded theory methodology to analyze data, various steps were followed in this study. As suggested by Charmaz (2006), the first step was initial coding, the pivotal link between data collection and the emergence of theory, and was intended to identify actions in each segment of information and assign concise labels that categorize and summarize data rather than applying preexisting categories to the data. In this coding stage, the researcher analyzed data line by line or incident by incident to find provisional and comparative codes that were grounded in data. At this point, several codes emerged and the researcher kept them short and specific.
Secondly, focused coding took place and tested the most significant or frequent initial codes by comparing them to extensive data. In this phase, memo-writing was useful to raise focused codes to conceptual categories. By writing memos, the researcher stopped and analyzed his ideas about codes and emergent categories as a means to formalize relationships into theoretical frameworks. Table 3 illustrates how data collected was initially narrowed to some preliminary codes. Table 3 Preliminary codes

Samples Preliminary codes
"Mis contribuciones en las discusiones grupales fueron malas porque soy el peor del grupo." (My contributions in group discussions were bad because I am the worst in the group.) "El trabajo en grupos me ayudó a mejorar el vocabulario y la forma de expresarme." (Working in groups helped me improve my vocabulary and how to express myself.) "En esta clase mis contribuciones fueron buenas ya que escuché y pude ser escuchada." (In this class my contributions were good as I listened and could be heard.) "Estar con compañeros que las lograban entender el vocabulario facilitó mi desarrollo en las actividades, y considero que mis aportes también llegaron a ser de ayuda para otros." (Being with colleagues who managed to understand the vocabulary facilitated my performance in the activities, and I consider that my contributions also helped others.) "Yo aportaba ideas y recopilaba las de los demás para así poder formar frases un poco más largas." (I brought ideas and compiled those of others so we could make longer sentences.) • Feeling frustration and negative self-image by contrasting others' high achievement and constant participation.
• Learning new vocabulary and expressing with ease.
• Socializing and exchanging knowledge and opinions.
• Progressing. Perceiving improvement little by little.
• Understanding the topic by working cooperatively.
• Improving the outcome of the task.
Thirdly, the axial coding phase helped the researcher to identify a core category or phenomenon by making connections between categories in the previous coding. In this stage, the researcher followed Creswell's (2012) instructions and selected one coding category, taking into account aspects such as its frequent appearance in data and its relationship to other categories, positioned it as the core phenomenon of the process under investigation, and related the other categories to it if they referred to causal conditions, strategies, contextual and intervening conditions, and consequences.
These grounded theory stages permitted gradual interpretation of those behaviors involved in the process of peer tutoring carried out by the participants of this study.

Validation
The qualitative process of data analysis is interpretive and requires the researcher's ability to report the findings. Thus, accuracy or credibility of the findings is of upmost relevance and is partly determined by the effectiveness of the instruments used during the implementation stage.
One strategy typically used by qualitative researchers to validate findings is triangulation. Creswell (2012) defines triangulation as the process of corroborating evidence from different sources, including participants, type of data, or tools to gather data. The purpose of triangulation is not to evince that different sources of data yield the same result, but to look for consistency in the relationship between the results and the phenomenon under investigation (Patton, 2015). Table 4 shows the triangulation among the instruments in this study that led to the identification of the categories. In this study, data collected was organized in an Excel sheet. Fragments of data from different instruments were constantly compared and hand analyzed in order to establish differences and similarities that yield future categories. Charmaz (2006) suggests three steps to analyze data: initial coding, focused coding and axial coding. Firstly, in the initial coding phase, the researcher analyzed data line by line and labeled emerging codes according to the relation between them. Secondly, focused coding occurred as a means for the researcher to develop and keep track of initial codes that were tentatively raised to categories. Inconsistencies that were found in this stage were not seen as weakening the validity of data but as an opportunity to synthesize and refine the initial organization of codes. Then, axial coding allowed the adoption of certain nascent categories as theoretical concepts.

Overall category mapping
The aim of this study was to assess the effects that peer tutoring and computer-mediated tasks had on students' spoken fluency. As a result of the data analysis process, three categories emerged from the coding stage that addressed the research question. These categories were: increased self-confidence when collaborating, enhanced perception of spoken fluency, and improved outcome of tasks. Figure 2 shows the relationship among focused codes that led to three categories. Figure 2. Category mapping. This figure illustrates the relation among subcategories and the emergence of the main categories and the core category.

Discussion of categories
In reporting findings, Patton (2015) recommends a balance between description and interpretation. Enough description and direct quotations are necessary to allow the reader to understand the situation and thoughts of participants represented in the report. Thus, substantial description is brought to discussion and organized in this study so that findings are manageable.
Similarly, the researcher intends to show the significance of results and whether patterns are clear, strongly supported by data, or just suggestive.

Increased self-confidence when collaborating
The analysis carried out by the researcher evinced a change in learners' confidence toward their learning thanks to the implementation of peer tutoring. The participants acknowledged the increase of self-confidence when collaborating in tasks. Initially, some learners with low achievement experienced frustration and negative self-image when contrasting their performance to others' high achievement and constant participation. Group organization that included tutors and tutees might have led tutees to think that others were superior and, hence, be reluctant to take active part in discussions due to others' reaction to their performances. Observe the following answers in learning logs after the first session: "Mis contribuciones en las discusiones grupales fueron malas porque soy el peor del grupo." (My contributions in group discussions were bad because I am the worst in the group.) (Excerpt, Student H, Learning log 1) "Mis aportes realmente fueron muy malos ya que no colaboré mucho en las discusiones." (My contributions were very bad as I did not collaborate at all in the discussions.) (Excerpt, Student K, Learning log 1) These excerpts show how students feel frustration regarding language acquisition when contrasting their learning progress with that of their classmates. The quality of others' interventions, including a broad bank of vocabulary, comprehending the task with ease, understanding teacher's instructions, correct pronunciation, openness to work with others, are some of the reasons for tutees to feel nervous and avoid taking risks in group collaboration. The following excerpt indicates that, in this moment, the students consider that they are to be taught by other learners and, thus, mainly assume roles as listeners.
"El trabajo en grupos me permite aprender sobre el inglés de los demás y que me enseñen a hablarlo un poco más." (Working in groups allows me to learn from others' English and that they teach me to speak a little more) (Excerpt, Student G, Learning log 2) Exposure to peer tutoring allowed learners to be aware of their leading roles in their own education. Early, in the third session, students assert that progress is perceived. As they advance, now they are confident to make their voices heard. Interacting with others becomes a two-way process as they exchange roles of listeners and speakers in a richer discussion and decisionmaking. The following excerpt demonstrates that, while peer tutoring was being implemented, participants evinced increased self-confidence.
"En esta clase mis contribuciones fueron buenas ya que escuché y pude ser escuchada." (In this class my contributions were good as I listened and could be heard.) (Excerpt, Student K, Learning log 3) Having the role of tutor can be perceived as a privilege. Accordingly, as self-confidence increases, one tutee reports the expectations concerning the scope of one's personal contributions during the process of peer tutoring.
"Estar con compañeros que lograban entender el vocabulario facilitó mi desarrollo en las actividades, y considero que mis aportes también llegaron a ser de ayuda para otros." (Being with colleagues who managed to understand the vocabulary facilitated my performance in the activities, and I consider that my contributions also helped others.) (Excerpt, Student D, Interview) Beginning this project, peer tutors were selected taking into account their outstanding performance in the needs analysis stage. An evidence of progression and improved selfconfidence is that tutees acknowledged that, although they lacked language mastery, they could help others. These students with low achievement experienced challenges in the development of the tasks, but they still considered their contributions could have benefited others. A teaching role arose almost naturally as well as willingness to help others, to convey and negotiate meaning, or to improve the outcome of the task. Regarding the initial tutoring role he/she assumes, one participant reports: "Me contribuyeron demasiado aquellos estudiantes de mayor nivel de inglés, ya que yo no tengo un buen nivel del idioma, pero puedo ayudar a los de menor nivel que yo." (Students with high level of English helped me too much, since I do not have a good level of the language, but I can help those with lower level.) (Excerpt, Student G, Interview) At the end of the implementation, most of the participants agreed that peer tutoring helped them to be more confident and comfortable when preparing to perform the tasks. At this stage, mistakes were not seen as frustrating but as a means to make progress. Differences in language proficiency were not remarkable.
"Aunque a veces me equivocaba, sabía que todos estábamos en el mismo nivel y que avanzamos juntos. Así que no hubo como ese miedo al qué dirán." (Although sometimes I was wrong, I knew that we were all at the same level and that we advanced together. So, there was no fear about what they will say.) (Excerpt, SB, Interview) Many of students' entries to logs provided enriching insights concerning their increased self-confidence during the implementation of this project. Therefore, in the interview, the participants were asked how self-confidence had influenced their performance in oral tasks. They responded that it led them to put aside embarrassment and perform the tasks with less pressure and more spontaneity that, consequently, fostered their spoken fluency. In agreement with this study, Nakata (2006) discusses self-confidence in relation to anxiety in language learning. According to the author, self-confidence entails a perception of communicative competence and concurrent low level of anxiety in using the second language.

Enhanced perception of spoken fluency
The second category that emerged from the analysis of data was the enhanced perception of spoken fluency. In connection with the previous category and the participants' intention to foster their fluency, one student reports an initial goal for the course.
"Poder hablar un poco más en inglés y no sólo unos segundos. Sin embargo, debo tener más confianza para no bloquearme a la hora de hacerlo frente a todos." (Be able to speak more in English and not just for a few seconds. However, I must be more confident so that I do not hesitate in front of everyone.) (Excerpt, Student B, Learning log 1) The rubric criteria used in this investigation allowed the researcher to measure, to a certain extent, whether the participants had progressed in their spoken fluency, considering learners' vocabulary and expressions, grammar, confidence and hesitation, ability to focus on the task, intonation and pronunciation. This analysis reflects the participants' spoken fluency before and after the implementation of this project, thus, students who did not manage to video-record the first and/or the final performance of tasks were not considered. Figures 3 and 4 show tutees' initial and final spoken fluency respectively.  "El trabajo en grupos me ayudó mucho porque nos colaborábamos y entendimos el vocabulario." (Working in groups helped me a lot since we worked collaboratively and understood the vocabulary.) (Excerpt, Student A, Interview) With new vocabulary, students felt that it was easier to make themselves understood in group discussion or when carrying out the tasks.
"Utilizamos las palabras que habíamos aprendido para dar a entender lo que pensamos." (We used the words we had learned in order to explain what we thought.) (Excerpt, Student K, Interview) As seen in figures 4 and 5, the following excerpt suggests that tutees' pronunciation was affected positively due to others' presentations.
"Nuestra pronunciación mejoraba ya que, si uno presentaba, otro entendía la pronunciación de palabras que desconocía." (Our pronunciation improved since if one students performed the task, the others understood the pronunciation of unknown words.) (Excerpt, SD, Interview) Conflicting with these statements, student G got a lower score in the last video in terms of vocabulary and confidence. This contradiction does not reflect stagnation; however, this might have been due to the different levels of complexity between the first and the last task. Therefore, although other tutees faced similar challenges in the last task, they managed to overcome difficulties more efficiently.
On the other hand, tutors' spoken fluency also benefited from assuming a teaching role.   "Se me dificultó el vocabulario ya que no tenía conocimiento de algunas palabras para organizar las ideas." (The vocabulary was difficult to me because I did not know some words to organize the ideas.) (Excerpt, Student B, Learning log 1) Perhaps unconsciously, one tutor expresses the relation between fluency and vocabulary by setting a goal for next session.
"Mis objetivos para la próxima clase son mejorar mi fluidez y ampliar mucho más mi vocabulario." (My objectives for next session are to improve my fluency and to increase my vocabulary.) (Excerpt, Student L, Learning log 3) Regarding this issue, Hilton (2008) states that lexical competence has a fundamental role in spoken fluency and, therefore, should be taken into account in language teaching.
Additionally, in the following excerpt, one tutor reports that pronunciation was affected positively due to others' contributions.
"En ocasiones también me ayudó mucho escuchar la pronunciación de mis compañeros para beneficiarme." (In some situations, I benefited a lot from hearing my colleagues' pronunciation.) (Excerpt, Student L, Interview) In some situations, tutors were asked to read texts as a means to motivate tutees to take active part in tasks. The following excerpt suggests that tutors also benefited from this exercise.
"Mi fluidez oral aumentó un poco por el hecho de leer textos. A medida que participaba, podía mejorar la pronunciación." (My oral fluency increased a little by reading texts. As I took part, I could improve pronunciation.) (Excerpt, Student B, Interview) "Percibo una mejora significativa en mi fluidez oral al leer textos." (I notice a meaningful improvement in my oral fluency by Reading texts.) (Excerpt, Student L, Interview) Generally, tutees and tutors' spoken fluency benefited from taking active part in tasks.
The constant comparisons of data between the different instruments used in this project evinced improvement due to the quality of the participants' interventions in more and more complex tasks.

Improved outcome of tasks
As it was stated in the previous categories, tutors and tutees benefited from group discussions in terms of confidence and spoken fluency. The third category arises since students improved the outcomes of the tasks in which they participated by committing to working in groups.
Developing trust with other learners might have led the participants of this study to set goals that are related to assuming compromise with the group. Students were acquainted that the responsibility for the results also depended on factors such as individual practice at home, discipline, willingness to cooperate with others, and personality (Sánchez, 2012).
After getting used to working in groups with new colleagues, one key factor that helped learners focus on tasks was comprehending what they were expected to do. One of the advantages of peer tutoring is that students receive more time for individualized learning and active helping and supporting (Topping, 2001). Peer tutors assisted tutees in this aspect since students being tutored could ask questions and receive repetition to ensure understanding. The following excerpts illustrate this assertion.
Similarly, the following excerpts suggest that peer tutoring helped learners increase the vocabulary used in tasks and, so, improve the quality of the tasks.
"En el trabajo en grupos, el vocabulario de unos complementa al de otros, haciendo mucho más rico el aprendizaje y la calidad del trabajo." (In group work, the vocabulary of some students complements others', enriching learning and the quality of tasks.) (Excerpt, Student L, Interview) "Hoy me causó dificultad hablar con personas que no conocía. Es más difícil hacer algo con desconocidos por la timidez. Al menos con un amigo el ambiente se vuelve agradable." (Today, it was difficult to me to talk to people I did not know. It is more difficult to do something with strangers because of shyness. At least with a friend the atmosphere becomes pleasant.) (Excerpt, Student B, Learning log 1) Alguna cosa que puedo hacer para mejorar el trabajo con mis compañeros es estudiar y comprometerme con el grupo. (Something I can do to improve work with my colleagues is to study and commit to working in groups.) (Student K, Learning log 1) "En las discusiones grupales de hoy yo trataba de aclarar las dudas de mis compañeros." (In group discussions today, I tried to answer the doubts of my colleagues.) (Excerpt, Student C, Learning log 3) "En la actividad de hoy tuve claridad a la hora de saber lo que estaba hablando." (In today's activity I clearly knew what I was talking about.) (Excerpt, Student D, Learning log 3) "El trabajo en grupos me ayudó a mejorar el vocabulario y la forma de expresarme." (Working in groups helped me improve my vocabulary and how to express myself.) (Excerpt, Student F, Learning log 2) In terms of enriching discussions and making decisions, one tutor reports that peer tutoring brought a wide variety of viewpoints that improved the results of tasks.
"Trabajar con otros amplió la gama de respuestas, puesto que cada persona tenía un punto de vista diferente, fue mayor la cantidad de ideas por lo mismo." (Working with others widened the range of responses, since each person had a different point of view, it brought a greater amount of ideas.) (Excerpt, Student L, Interview) By taking part in group discussions, learners could elaborate their ideas and add complexity to tasks.
"Yo aportaba ideas y recopilaba las de los demás para así poder formar frases un poco mejor elaboradas." (I brought ideas and compiled those of others so we could make thorough sentences.) (Excerpt, Student B, Learning log 3) Peer tutoring was effective and meaningful for students as a result of the participants own attempts to approach working in groups (Oxford, 2003).

Core category
Given (2008) defines a core category as "the main theme, storyline, or process that subsumes and integrates all lower level categories in grounded theory, encapsulates the data efficiently at the most abstract level, and is the category with the strongest explanatory power" (p. 131). The teacher-researcher in this study could establish a core category after going through the process derived from the initial, focused and axial coding stages. This category appeared frequently in data and was strongly related to other categories. Creswell (2012) suggests that the core category is positioned as the central phenomenon under investigation and relate to it other categories that denote causal conditions, strategies, contextual and intervening conditions, and consequences. Thus, the interrelationship was portrayed in Figure 7 using a coding paradigm. Figure 7. Coding paradigm. This figure illustrates the emergence of the core category.

Conclusion
In summary, this study has yielded meaningful recommendations regarding the phenomenon of learners' lack of spoken fluency in L2. Participants involved in peer tutoring and computer-mediated tasks evinced increased self-confidence that allowed them to perform tasks more naturally and enrich discussions by adopting roles of tutors and tutees. Progress in spoken fluency was perceived due to collaboration and learners' willingness to improve the outcomes of tasks. These results suggest that language teachers can make the most of differences in learners' achievements by encouraging them to assume the prestigious roles of tutors.

Introduction
Throughout this project, illustration has been made of the procedures carried out to address ninth grade learners' difficulties to speak fluently in L2. Peer tutoring and computermediated tasks encouraged students in three important aspects of learning: increased selfconfidence when collaborating, enhanced perception of spoken fluency and willingness to improve the outcomes of tasks. The results of this study convey implications for current pedagogy and future research interventions regarding payoffs and pitfalls found, specifically in relation to spoken fluency and their significance to the EFL Colombian context.

Comparison of results with previous studies' results
In the present study, the researcher sought to identify the effects of peer tutoring on twelve ninth graders' L2 spoken fluency when performing computer-mediated tasks. No previous research that combined these constructs to address young learners' lack of spoken fluency was found. Data was gathered with learning logs, video recordings, rubric to assess spoken fluency and a final semi-structured interview. Data analysis suggests that the implementation of this project affected the participants in terms of spoken fluency, selfconfidence when collaborating and their willingness to improve the outcome of tasks. These results reaffirm, validate and expand the findings in other investigations. (2016)  Usma (2015) studied the effects of interactional tasks (e.g. role plays, information gap activities) on eleven administrative workers at a private university to address their lack of spoken fluency. The results showed that the participants recognized the importance of fluency and that the strategy proposed had brought positive outcomes. Likewise, the present study indicates that learners' spoken fluency is fostered by interacting in tasks, and this assertion expanded to online environments.

Doman and Bidal
Regarding the use of tasks and online learning environments to foster spoken fluency, Castiblanco (2014) investigated the effects of Task-fluency Discussions in Second Life on eight A2 university students. The participants needed to improve oral fluency and gain more motivation in virtual learning spaces since they were required to take part in a blended course.
The results suggested that students increased awareness of learning strategies, self-effort to speak fluently, and self-perception of being a fluent language user. Similarly, the present study reports that oral tasks foster spoken fluency when they are held using an appealing online environment.

Significance of the results
This action research study exalts the significance of using peer tutoring in EFL teaching and its contributions in terms of spoken fluency and oral tasks in online contexts.
For the researcher, this investigation represents the satisfying outcome that could be developed thanks to the rapport provided by previous studies and authors whose contributions in the fields of peer tutoring, task-based learning, computer-mediated instructions and spoken fluency are valuable and remain valid for language teaching. This research study opened a new vision for the improvement of curriculum, lesson planning and classroom management that, in the case of the researcher's teaching practice, will keep on developing in order to provide learners with meaningful interactions in L2 where they can collaborate confidently, speak spontaneously and be aware of their learning processes.
For the participants in this study, being involved in peer tutoring allowed them to assume coaching roles that, consequently, increased collaboration, enhanced spoken fluency and their willingness to improve academic results. Additionally, peer tutoring fostered interpersonal relations since learners reported that, beginning the project, it was difficult to work with people they did not know. Later, it was evinced that they became friends and reinforced bonds.
Learners' openness to accept others' points of views and positive attitudes toward collaboration were paramount for this project to succeed.
In Colombia, peer tutoring needed, and still needs, further exploration, especially with regard to its effects on young L2 learners' abilities where little research has been provided. Thus, the findings in the present study are of highest importance since they become a starting point for developing politics that could address learners' difficulties. The results become a great prospect to involve teachers in a pedagogical research and knowledge upgrading by transferring conclusions to similar contexts, subjects, skills, among others.

Pedagogical challenges and recommendations
One of the most salient behaviors that hindered students' speaking performances was their initial reluctance to work in groups with strangers. In this regard, it is recommended to choose match activities that promote friendship and trust so that students establish stronger bonds within the learning environment. However, when a friendly atmosphere among learners is built, especially in same-age tutoring, teachers must ensure that some students do not go off task as they are working with their new friends. Future research should consider the steps Carreño (2014) suggests for establishing an online learning community in which teachers and students' abilities come to light.
Regarding the use of tasks, the ascending degree of difficulty suggested in this research study might have hindered some of the students' progress. Even though the intervention yielded positive outcomes, in one particular case, anxiety and hesitation were highly manifested. For future researchers willing to replicate this study, it is recommended that the difficulty of tasks agrees with learners' abilities to perform them naturally.
Furthermore, peer tutoring can a be time-consuming strategy for teachers as well as an additional workload for students. Since learners already have the responsibility of their own learning, researchers must consider a plan for alternating roles so that all participants can train and be trained.
Peer tutoring can be a powerful strategy to extend the influence of instructions and increased opportunities for students to respond in small groups (Mitchell, 2004). One special recommendation is that teachers do not forget that, even when tutors and tutees are collaborating in groups, positive student-teacher relationship needs to keep on developing. That is, teachers must be present for immediate support as they have guiding, motivating, inspiring, monitoring, and many other roles in classrooms.

Research limitations of the present study
This study presented some limitations being the most noticeable one the scheduled timeline. Due to curriculum and teachers' greater concentration on formal aspects of language, to the detriment of real communication in L2, the intervention of this research intended to bring an additional space for students to interact orally in English. It was held using an online collaborative classroom that allowed the participants to attend the sessions from home. Although most participants joined the meetings, sometimes it was difficult to schedule a frame that matched students' availability.
The use of a computer-based classroom represented a limitation in one specific aspect in this investigation. The participants had not attended online lessons before this project. However, the implementation was carried out with no major difficulties. Students were required to download an application in case they tried to access the sessions using their mobile phones.
Regarding this aspect, one participant could not join the meetings due to incompatibility between his device and the software. Then, in using computers, tablets or mobile phones, it is advisable that teachers anticipate software compatibility or learners' comprehension about how to access the sessions in order to guarantee their participation.

Further research
This study focused on enhancing students' spoken fluency through peer tutoring in computer-mediated tasks. The results revealed increased self-confidence in learners that allowed them to collaborate in the activities, enhanced perception of spoken fluency and additional effort to improve the outcomes of tasks. It would be worth that future research investigates the effects of peer tutoring on other skills, or to explore the effects of other types of peer tutoring such as cross-age tutoring with more participants or an increased number of interventions. The researchers must bear in mind that, in peer tutoring, it is paramount to monitor mentors and mentees' performances in order to consider timely adjustments and motivate mentees to assume coaching roles.
An interesting research opportunity is open for researchers willing to replicate, validate this study or to generalize the findings, specifically regarding the use of tasks to improve spoken fluency. Future research should have into account the complexity of tasks as, for instance, a decreasing level of difficulty in the activities will yield unreliable high results. In this sense, Robinson (2011b) recommends that tasks are sequenced for learners from simple to complex so that they approximate the demands of real-world needs. Varying the complexity of tasks affects language acquisition and learners' performances outside the language classroom.
Additionally, future research must consider some conditions for fluency development.
Hinkel (2017) states that fluency is related to spontaneity, thus, in measuring its progress, researchers must comply with certain guidelines that guarantee natural performances. First, students must be familiar with the material worked, that is, no unfamiliar vocabulary or grammatical constructions. Second, the focus of tasks should be on conveying meaning. Third, the participants should be encouraged to perform tasks at a faster than usual speed. Fourth, there should be quantity of practice.
In using computers, tablets or mobile phones to congregate learners, it is advisable that the participants have clear instructions concerning expected behaviors, safety in the Internet and its resources, ethical considerations, and how misconducts are going to be tackled (Provenzo, teachers agree with specific rules in an Acceptable Use Policy (see Appendix D).
Although future research focuses on exploring the effects of any strategy on students' spoken fluency, it is recommended that the four strands suggested by Nation (2007) are devoted equal class time and learners' efforts given that the development of any skill might contribute to the others.

Conclusion
This chapter illustrated final insights from the study of the effects of peer tutoring in ninth grade learners' spoken fluency. The analysis of data collected highlighted three benefits of the intervention of this project: increased self-confidence, enhanced spoken fluency, improved outcomes of tasks. These findings suggest that the initial differences in learners' achievements were not a weakness, but they became the means for students to accomplish the goal of their learning.
The combination of peer tutoring and computer-mediated tasks emerged as a means to provide learners with an additional space to interact orally. Lots of work in the ESL field permitted the researcher to identify a gap in research, especially at the local level. Since no previous exploration that addressed young learners' spoken fluency in L2 through the use of the selected strategies was found, this study represents a starting point for future research willing to transfer findings to different settings. Thus, this inquiry was possible thanks to authors whose contributions and guidance to second language learning and teaching served as rapport. This investigation is useful for acknowledging and emphasizing on the importance of helping learners speak naturally, foster language proficiency by instructing others, and improve affective factors such as motivation and confidence.
Finalmente, es de aclarar que este proyecto hará uso de dispositivos electrónicos como computadores, tabletas o celulares inteligentes para la realización de las actividades. Por esto, es necesario que todos los involucrados estén familiarizados con el documento Políticas de Uso • Contrast different points of view. • Present evidence to justify answers. • Distinguish language needed to foster discussion.

LEARNING OUTCOMES
By the end of the lesson, the learners will be able to: • Express advantages and disadvantages of mobile phones.
• Use specific vocabulary to talk about good and bad habits when using mobile phones.
• Give advice about how to improve relationships with relatives and friends.

ACTIVITIES Time
Pre-task: General overview of the task and vocabulary. Learners see some pictures about people using their mobile phones in different situations. Students vote for or against people's behaviors in the pictures (Annex 1). Students might justify their choices.
During task: Using their mobile phones or tablets, students will take a survey called "Are you addicted to mobile phones?" (Annex 2). The teacher will check students' results and interpret them (Annex 3). Students will explain the results of the survey and discuss whether or not they are addicted to mobile phones. Survey available at: https://goo.gl/forms/ky1snYy7aTZuh9wg2 Students classify into two groups (good and bad habits) the activities they found in the survey. Students justify their choices, express their opinions and defend a point of view.
Students design a poster proposing solutions to the problem of lack of communication with relatives and friends and present it to the class.
Post-task: Using learning logs, students reflect on their performances during the tasks.