Interweaving Autonomous Learning and Peer-tutoring
in Coaching EFL Student-Teachers^*

La conexión entre aprendizaje autónomo y la tutoría entre pares como apoyo a
estudiantes de inglés como lengua extranjera

In this article we share the findings on the exploration of undergraduate efl (English as a Foreign Language) student-teachers' professional preparation and autonomous practices. Participants were tutees who attended tutorial sessions with peers in higher semesters. In this context, tutoring was based on a model we designed. As the model was implemented, they collected information by means of multiple qualitative research instruments. Results indicate that tutees' learning process was strongly influenced by their personality and attitudes. Similarly, tutees expanded their views in regard to the English language, its learning, and their communicative competence. Finally, we pinpoint some considerations for others interested in adopting this pedagogical strategy.

Key words: Peer tutoring, university tutoring, autonomous learning, independent work

En este artículo compartimos hallazgos relacionados con la preparación profesional y las prácticas autónomas de estudiantes de inglés en un programa de licenciatura. Los participantes tutorados asistían a sesiones tutoriales con compañeros en semestres superiores. En este contexto, la tutoría se basó en un modelo que diseñamos. Durante la implementación del modelo se recogió información a través de múltiples instrumentos cualitativos de investigación. Los resultados revelan que el proceso de aprendizaje de los tutorados se vio fuertemente influenciado por su personalidad y sus actitudes. Asimismo, los tutorados ampliaron sus puntos de vista sobre la lengua inglesa, su aprendizaje y su competencia comunicativa. Finalmente, señalamos algunas consideraciones para los interesados en adoptar esta estrategia pedagógica

Palabras clave: Tutoría entre compañeros, tutoría a nivel universitario, aprendizaje autónomo, trabajo Independiente

The initial idea to undertake this research process is rooted in various issues. To begin with, our students who were enrolled in the Modern Languages Program at a public university evidenced limitations in their English proficiency. A second concern referred to national educational policies at the tertiary level based on decree 2566, issued in 2003 by The Colombian Ministry of Education (MEN), which suggested the adoption of the credit system. A credit was defined in relation to the time students devoted for their independent work. Thus, this policy led universities such as ours to open supportive spaces which fostered students' autonomous work by means of tutoring. In this sense, it was expected that students could become more efficient in managing their academic time and optimize their learning process. In our university, professors started to provide this tutoring to their students. Unfortunately, we lacked clear guidelines to support learners by means of this strategy. Hence, our study aimed at understanding the relationship among tutoring, tutees' professional development and their autonomous practices.

Bearing the previous considerations in mind, we embarked on an action-research journey. First, a diagnostic stage to locate the sources of problematic situations was conducted for about one academic term. By means of an initial survey applied to first and second semester students as well as professors in the language courses, the analysis of tutoring experiences as debutant professors, and the revision of the existent documents the university had produced about tutorials, we established the concepts this population had about autonomy and tutoring. The information collected in the diagnostic stage revealed that participants perceived tutorials as personalized spaces to get support through feedback. With regard to students' perceptions about autonomy, we discovered that they conceived autonomy as an independent practice carried out in their free time guided by their initiative to make decisions. Additionally, self-confidence and responsibility emerged as key features in facing personal limitations. The most remarkable hindrances behind tutorial development were that both professors and students had limited time for these meetings and only a few students attended tutorials due to their lack of confidence with educators.                 As a consequence of the former evidence, we were encouraged to plan tentative tutorial sessions. Initially we decided that peer-tutoring would be the most suitable type of support for participants. We also decided to form an emerging research group with student-tutors of higher semesters. Over a period of four weeks tutorial sessions were held and explored. Information from tutors' journals, interviews with tutors and tutees, audio and video recordings of sessions and tutorial description formats guided us to propose a threecomponent peer-tutoring model, suitable for the context of our research. This model is composed of the basic principles to conduct tutoring sessions, the tutoring cycle and the ideal tutor's profile.

The tutorial model we previously referred to was implemented for an academic year. Thus, we could answer our peer-tutoring central query, which aimed at exploring what features of first semester Modern Languages students' professional preparation emerged as they took part in an autonomous-learning based peertutoring model. To tackle this question we formulated three objectives. On the one hand, we sought to determine the role that the peer-tutoring model had in tutees' language development. Secondly, we explored how tutees' participation in the model revealed socio-affective and teaching concerns. Lastly, we examined the evidence of tutees' autonomous practices revealed along their participation in the tutorial meetings.

In the forthcoming sections, besides providing more details about the investigation process we undertook during the first phase, we will discuss the main pillars underlying this project from both theory and research experiences. Moreover, we will include a description of the research design we implemented to answer the question, which constituted the second stage of this study. Additionally, we will reveal the findings followed by our conclusions and pedagogical implications.

This literature review looks at several concepts related to peer-tutoring and autonomy. Starting from these considerations, the principles of these two approaches will be discussed. Finally, relevant research experiences in regard to the areas being studied will be commented on.

Peer-Tutoring Principles in the University Context

Revising literature in the field, several concepts such as tutoring, peer learning and tandem can be useful to define peer-tutoring. In general tutoring implies supporting and monitoring students constantly so that they can achieve their integral preparation during their studies. Within the previous framework, Álvarez (2004) define tutoring as guiding and advising pupils. Peer tutoring emerges as a specific category of tutoring in which two students, a tutor and a tutee, are involved. The tutor does not necessarily have to be an expert or a teacher in the target area since it is expected that his guidance will lead the tutee to find answers and solutions by himself (Topping, 2002).

Peer-tutoring bears certain similarities with peer-learning (Sampson & Cohen, 2001) and tandem (Stickler 2006); these approaches refer to the support that peers offer their equals in order to achieve learning goals. Thus, several social and psychological factors around these experiences are alike. However, in tandem and peer-learning, peers mutually support each other's knowledge building, whereas peer-tutoring implies one of the two becoming a permanent guide.

Peer-tutoring in the context of our study refers to a personalized extra-class space where a student guides another one in order to support him or her in his or her integral preparation (Viáfara & Ariza, 2008a). Taking into account students' status in peer-tutoring, Gaustad (1993) has distinguished between peer-tutoring –in which peers are of similar age and academic status– and the ones in which the tutor might be older or more competent in the target area than the tutee.

With regard to the organization of peertutoring, the following aspects were considered to design the model for this study (Viáfara & Ariza, 2008a). We started with a needs analysis to set suitable goals for the program (Sampson & Cohen, 2001; Álvarez, 2004). Then, careful planning of the logistics was required to arrange schedules and spaces, set pairs together and locate resources (Topping, 2002; Sampson & Cohen, 2001; Álvarez, 2004; Stickler, 2006). Finally, as Sampson & Cohen (2001) state, managing the process by introducing the program to participants, supporting them constantly and evaluating the tutoring experience strengthened our enterprise.

Academic and Social Implications of Peer Tutoring: Gauging Research Experiences

Studies in relation to peer-tutoring have revealed valuable information in the last years. Álvarez & González (2005) determined that peer-tutoring reduces the distance between participants since they are both students. This positive environment stimulates learning because they perceive each other as being part of the same group. In addition, as Mynard & Almarzouqui (2006) state, cooperation fosters a safe atmosphere where asking questions, solving doubts and discussing topics become easier. In regard to academic benefits, peer-tutoring favors at risk students because it promotes the practice of specific communicative and social skills. Luca, Cowan & Clarkson (2003) highlight the role of peer-tutoring as a stimulator of critical thinking and, in general, cognitive processes. Moreover, this type of experiences increases university program retention (Torres, 2002; Packham & Miller, 2002). Embracing learning strategies and study skills has also been pinpointed by Beasley (1997) and Mynard & Almarzouqui (2006) as gains in this kind of experience. All the same, these spaces foster participants' leadership and make both tutors and tutees more competent in regard to their instructional potentiality (Álvarez & González, 2005). Likewise, Torres (2002) underlines how peertutoring allows tutees to become more prepared when participating in classes and handling extra academic issues.

Conversely, investigators have detected some hindrances in peer-tutoring. The lack of administrative support and logistics problems regarding flexibility in schedules as well as spaces have a negative impact on tutees' attendance to sessions (Álvarez, 2002). On the other hand, tutees' dependence on tutors and some tutors' lack of preparation may cause a mismatch between their expectations about tutorial development (Mynard & Almarzouqui 2006; Beasley, 1997).

Autonomy as a Corner Stone in Peer Tutoring: Concepts and Conditions

Peer-tutoring experiences imply that tutees gradually move from being supported by someone with higher expertise to become their own tutors. Thus, as Suárez (2002) mentions, there is a connection between tutoring and autonomous learning. Expanding on the previous idea, autonomy is a concept that has been explored from many different perspectives. From philosophical and psychological perspectives, it means becoming responsible members of a social group, whereas in the political field, autonomy implies freedom from external control. At the educational level, there may be a risk of implementing autonomous approaches which favor a particular function of instruction. Coleman (1996) emphasizes the threat of implementing autonomous hegemonies that deny individuals the possibility to develop their own meanings of education and their autonomous practices. As posed by Riley (1988, p. 20), "learning is a social process and varies according to the nature of the society a learner is part of ". Autonomous practices are highly determined by the cultural patterns of a specific social group.

As we explored the concept of autonomy, we learned that Holec (1981) refers to this paradigm as a set of abilities an individual possesses to take more responsibility in various social matters. Within this general framework, autonomy is conceived from narrow positions in which learners may face their process in isolation to social versions of autonomy in which assisted learning is the core.

In the same spirit, autonomy can be understood as "the capacity students have for detachment, critical reflection, decision making and independent action" (Little, 1991, p. 4). Low (1996) and Dickinson (1995) refer to autonomy as the ability and attitude learners apply to take charge of their own education. To our understanding, being autonomous implies optimizing a viewpoint towards knowledge construction to decide what to learn, when and where learning should be developed, materials to be used and ways to assess one's process.

Little (2006) and Vieira (2002) mention some attitudes and conditions which favored autonomy as one's willingness to look individually and collectively at educational contexts from a critical perspective. This position may shift from negotiation to confrontation of authority. The previous implies that pondering our actions is essential to know oneself and to understand one's role in education showing tolerance with uncertainty and ambiguity. Moreover, another condition which affects autonomous practices might be the existence of suitable environments for students to master the language through its use.

The previous conditions can be considered when moving to the practical implication of autonomy in educational settings. To begin with, when trying to create autonomous learning environments, Scharle and Szabó (2000) and Vieira (2002) proposed an awareness that raises dynamics which may include attentiveness in learning how to learn and the social nature of language to foster communicative competence development. Another stage is fueled by learners' changing attitude towards their process, their roles and the factors involved in it. As a result, learners might face the challenge of transferring roles within class settings.

Revising Research on Autonomous Practices and Profiles

Among recent research reports in this area, Luna & Sánchez (2005) identified four learners' autonomous profiles. The first one includes students who depend on others' initiative about what and how to work. A second one refers to students who reveal certain attitudes of taking charge of their own learning outside the educational context. In third place students showing initial reflection towards autonomous learning are described. Finally, they mentioned students who, by means of a high determination to take decisions, seek to reach their objectives.

Within the context of ALEX, a language learning program based on a self-access center at Universidad Nacional, Lagos & Ruiz (2007) explored some of the principal factors that affected the exercise of autonomy in this learning community. Among these factors, they revealed participants' conception of autonomy and their previous experiences. Shortcomings such as the lack of programs to foster autonomous learning, limitations in logistics support and participants' attitudes were detected in the exercise and development of autonomy.

In the same spirit, Ariza (2008) reports a study which aimed at identifying EFL students' understandings of autonomy evidenced through their learning in and outside the educational setting. Students' conceptions of autonomy were linked to learning beyond the classroom and facing a constant dilemma between detachment and teacher dependence. This situation generated feelings of satisfaction, frustration and uneasiness in participants. Another core issue was students' independence when making decisions supported in continuous reflection. This dynamic fosters participants' learning through experience and permanent evaluation processes.

Studying the literature and what researchers have previously shared in the field enlightened us in the observation of what happened when the peer-tutoring model was implemented in our setting. Bearing the previous in mind, we will continue with sharing how peer-tutoring took place while it was investigated.

As we stated in the introduction, by means of a first action research cycle, we studied our context to propose a tutorial model (Viáfara & Ariza, 2008a). The second stage of our research, which is described in this article, focused on studying what features of first semester Modern Languages students' professional development emerged as they took part in an autonomous-learning based peer-tutoring model.

Study Scenario

Our Program, Modern Languages, prepares students for the teaching of English at secondary levels. It includes seven five hour-weekly English language courses, called Communicative Projects, to support pre-service teachers' foreign language communicative competence. The methodologies implemented to guide students' learning are basically the Communicative Approach and Project Work. Since class time is reduced, students are expected to work independently to achieve the aims of courses. Students also take grammar and phonetics classes. A course of Applied Linguistics, three courses of Didactics as well as the Teaching Practicum, which constitute the pedagogical block, are expected to foster reflection and research.

The program syllabus is supported by tutoring which aims at preparing students in their courses and is generally offered by teachers in their regular working schedule. These professors generally help students individually providing two tutoring hours a week per course. Students attend these sessions to get support on specific tasks, in homework revision and sometimes to be evaluated.

Based on the experience gained during the diagnosis and piloting stages in 2006, during 2007 first and second semester students were invited to participate in the project. We agreed to continue offering tutorials to several former participants from third and fourth semesters. We expanded tutorial time from one hour to two hours weekly for former tutees. Finally, sixth semester students enlarged the group of tutors.

In relation to tutors, we continued a rigorous selection process to ensure that they had an integral profile to undertake their role. They submitted curricula vitae and had an interview. Once the group was formed, the main researchers focused on organizing participants in pairs (tutor-tutee) based on their time availability and spaces accessibility.

Along the first and second semesters in 2007, the tutorials were implemented in the Modern Languages Program. Tables No. 1 and 2 summarize relevant information about the tutors and tutees that participated in the process.

Tutors' Preparation and Implementation of the Model

Tutors' preparation involved their understanding of the tutoring model and the principles behind it. The first component of the model is rooted in three essential principles. To begin with, peertutoring session sought to support tutees in their English proficiency level. Along with this academic objective, students might get support in aspects concerning their integral preparation as future EFL teachers. The last principle was focused on empowering tutees as learners by means of strategies in learning how to learn.

The second component of the model is conceived as a cycle (See Figure 1). It describes the ideal process that might orient tutorial sessions development. A tutorial session is nourished from tutees' immediate needs or their pre-established plans for long term improvement. During the sessions, tutors and tutees may explore and practice specific language aspects while feedback is constantly provided. The tutorial session is expected to encourage tutees' independent work. A new cycle will start when self and peer evaluation provides information that would reorient the tutoring process.

The profile of the tutor emerges as the last component in the tutorial model. As it supports a pre-service teacher preparation program, the tutor needs to display values, among them responsibility and respect as well as good communication skills, in order to keep a harmonious relation with the tutees. Likewise, being enthusiastic about the foreign language and its teaching becomes essential. Additionally tutors' skills in providing information about material sources and learning strategies support tutees' autonomous learning.

As evidenced in the center of Figure 1, we incorporated different tools within a reflective approach to guide tutors in handling the model. The richness of reflection through investigation strengthened the tutors' learning when they reported research findings, shared journals and attended academic events. We provided opportunities for real tutorial session video analysis and tutorial simulations which informed newcomers about real situations of interest. Problem solving tasks involved critical incident analysis in which expert tutors' experiences coached newcomers. Additionally, reading about important theoretical principles underpinning tutoring and autonomy was central to lead participants in developing skills to connect theory and practice.

In 2007 we simultaneously implemented the peer-tutoring model and conducted our investigation. Qualitative research under an action paradigm was the base for our study. "The aim of qualitative approaches is to offer descriptions, interpretations and clarifications of naturalistic social contexts" (Burns, 1999, p. 22). On the other hand, action research is conceived as "on-the-spot procedure designed to deal with a concrete problem located in an immediate situation" (Cohen & Manion, 1994, p. 192). Thus, the starting point of this cyclical methodology is a diagnostic stage which leads to the planning and implementation of various strategies. While the action is in progress, constant reflection fueled by observation and other research techniques takes place. Finally exploring the impact of strategies should guide investigators to start a new cycle.

Framed by this qualitative paradigm, primary and secondary instruments were used to collect evidence and achieve the objectives our study sought. By means of semi-structured interviews conducted at the end of each term, we elicited participants' views about their experiences in the tutoring program. We guided the interviews on a pre-established set of questions (See Appendix 1a, 1b). The two main researchers, with the help of some tutors, conducted 48 interviews for tutees and 24 for tutors along 2007.

Secondly, tutors' journals provided substantial information since they described tutees' attitudes, concerns, and actions during the tutorials. When tutors became members of the research group, we guided them in what a journal was and how to keep it in relation to our research focus. We all agreed on writing one entry after every tutorial session or once a week. Thirteen journals were analyzed by the end of the first semester and 12 in the second term.

Finally, tutorial sessions were audio and video recorded. We aimed at observing how participants behaved while they were engaged in the tutoring model. We agreed on a specific schedule to record the sessions, depending on logistics. During the first semester in 2007, 33 sessions were audio or video recorded, while in the second term 10 sessions were registered.

Additionally we also counted on secondary instruments. To begin with, 168 tutorial sessions in the first semester, 2007, and 106 in the second term were described in the formats that the research group had designed. Along with information to specify the participants' identity, time and place of tutorials, the formats gathered descriptive data about the tutees' needs, objectives, achievements, methodology of work, materials and suggestions. Another secondary instrument was the minutes of the research group meetings, which contained records of discussions about tutorial development.

Following Burns' suggestions for data analysis (1999), our process included assembling, coding, comparing and building interpretations of data. We started by the exploration of tutors and tutees' interviews to find common patterns to answer our question. Individually we used color coding procedures to identify emerging features. We proceeded similarly in terms of journals, audio and video recordings. Simultaneously, we used information from secondary instruments to validate what we had found. Then, we confronted our embryonic findings. Investigator triangulation gave weight to our findings. Similarly, data triangulation was also evident as multiple sources of information were used (Janesick, 1994). Furthermore, as another strategy for internal validity (Merriam, 1988), we collected data for a long period of time over two consecutive academic terms.

As a result of this first stage, 25 outstanding issues were identified as we related information by means of comparison and contrast. A second stage was guided by our research objectives which enlightened us to reduce and classify the patterns detected. This cycle continued until we established three main categories with their corresponding sub-categories which are fully discussed in the coming section.

Initially our research project's main focus was examining the relationship among peer tutoring, autonomous work and tutees' English proficiency level; however, as we dug deeper into the phenomenon, we realized that other issues were interwoven. Thus, we expanded the scope of our study to include three dimensions which we identified as a core for tutees' integral preparation as future EFL teachers. The first feature is tightly connected to language learning and has been denominated, "Gaining Experiential Knowledge in Reaching and Envisaging Language Learning Goals". The second issue involves socio-affective factors discussed under the heading of "Brick by Brick: Building up Confidence". Lastly, a core principle in autonomous learning, decisionmaking guides the discussion under the title, "From Engaging to Acting: Shapes of Tutees' Willingness to Undertake Their Learning".

Gaining Experiential Knowledge in Reaching and Envisaging Language Learning Goals

This first trait correlates with our intention to determine the role that the peer tutoring model had in tutees' academic profile. In this line, data revealed that the primary drive for tutees to become involved in tutoring was to fulfill their language needs. They succeeded in broadening their gains in relation to their empowerment as learners as also reported by Beasley (1997), Mynard & Almarzouqui (2006), and Velandia (2007). Additionally, tutees reflected upon themselves as future tutors or teachers which matches Álvarez & González (2005), who highlight the instructional potentiality participants can gain from tutoring.

Along peer tutoring, tutors became aware of tutees' styles and showed concern about helping them identify these preferences. Thus, they counted on valuable information to support learners in their language improvement as much as to provide them with guidelines for sessions development. Tutees acknowledged this effort of their tutors and they considered that their individual characteristics were respected.

Several participants mentioned how taking into consideration tutees' learning styles contributed both to their skill development and comfort; D, a tutor, expresses:

Planning Strategies to suit Tutees' Needs

The previous consideration of tutees' presumable learning preferences led participants to explore their needs as a diagnostic stage in planning the tutorial spaces. In regard to reported limitations, language aspects and their incidence in tutees' performance in classes appeared. These shortcomings originated from pupils' eagerness to tackle many language aspects simultaneously, their high school fossilization and low proficiency since they lacked opportunities for language use in classrooms. Needs were explored when tutors and tutees reflected during their examination of previous knowledge, their learning process. They also reflected on themselves as persons. Additionally, they engaged in self and peer evaluation which provided feedback and fostered their selfmonitoring. The following excerpt from a tutorial session audio-recording evidences the dynamics described above:

Identifying tutees' needs preceded proposing a working agenda. This agenda involved a good amount of real life topics and situations and was sometimes set by tutors or by the two of them. Organizing this plan implied using various sources, techniques and strategies for learning such as videos, games, grammar exercises, detailed explanations, skill-based work, homework and dictation. This selection responded to tutees' styles and proficiency levels. The previous route flowed into the tutees, achieving a range of goals. In relation to their immediate needs, tutors and tutees informed that tuteesr clarified doubts and corrected many of their problems.

Figure 2 below summarizes specific findings in relation to tutees' main academic gains as they participated in tutorial sessions. A first area of the profits involved issues concerning language proficiency and understanding how English works, as the following sample reveals. "When I was listening to her, I almost couldn't believe she was talking so fluent and with very few mistakes, I think she is making progress" (J⁴ Tu 1 2007).

Closely connected with the previous gain, participants reported how incorporating certain study skills and strategies made them better not only at communicating through the language, but also at changing their learning styles for more favorable ones as evidenced in the following excerpt when a tutor was asked about her tutees' gains: "They gained not only in fluency but in taking risks in using English for communicating; to feel more secure so they feel they are able to communicate" (I Tu II 2007). In the same line, another aspect was identified regarding the changing nature of tutees' beliefs. They realized that language learning is more than translating, grammar practice or memorization. In these sessions language was used for real purposes while the classroom was regarded as a space to be provided with input. Likewise, in the excerpt below, tutees reflected on the connection between tutorials and autonomy "... To be aware of one's willingness to reach a goal by one's own means" (I Te 1 2007).

This autonomous trait was also perceived in tutees' transferring of their abilities to other classes taught in English as the following tutor revealed in an interview: "Along the process my tutees told me that when they had to work on readings in English classes with their professors and they came across those words we studied during tutorials, they tried to remember how I told them to pronounce them". They also shaped their attitudes and behaviors in order to build up character in facing learning and life in general as expressed by a tutee: "looking at the tutors' proficiency to speak, I wonder when I would reach the same level... you know, and one realizes that what we do in classes and tutoring is not enough; one has to be self-demanding to achieve one's goals" (I Te 1 2007). Though the general perception of tutees in regard to tutorial sessions was positive, tensions of a different nature emerged. To begin with, there were cases in which tutees questioned tutors' knowledge of the language and procedure as well as the selection of relevant issues in the sessions as they claimed they trusted teachers more.

The previous was not the only mismatch between tutors and pupils' expectations; others referred to aspects concerning the logistics such as tutees inviting another peer to the session or the extent to which a tutor had to support them in their duties.

Brick by Brick: Building up Confidence

Though tutees' main objectives ran towards English language practice and development and the socio-affective dimension that tutoring nourished became a pillar in what they gained. Specifically, confidence was identified as the core building block around which several issues and opportunities were shared so that they grew as learners and future teachers. Similar findings have been reported by Álvarez & González, 2005; Mynard & Almarzouqui, 2006 and Packham & Miller, 2002.

In spite of the fact that tutors were simultaneously teachers and learning peers, being partners prevailed while working together. The previous considerations implied that the relation between the two actors in tutoring was less tense than with professors and more open to multiple options as to how tutors and tutees became acquainted. When tutorials started, feelings of uneasiness were common since tutees thought their tutors would judge them because of their limitations. However, they came across a working dynamic in which formal assessment was not used. The previous circumstances fostered a non-threatening atmosphere free from criticism and full of encouragement.

Bearing the previous in mind, many testimonies of tutees expressing how they reduced their shyness, increased their self-esteem and selfassurance were found. An outstanding situation we witnessed involved a participant who belong to the project for about one year and a half. Similar to many students from these regions who are of rural or semi-rural origins, he was characterized as being so shy that he would not speak to most people around or even keep eye contact while communicating. In the following evidence, this participants' tutor commented about his progress:

Student-tutors allowed a closer relationship with their pupils by accepting and promoting tutees' free expression and being receptive to their common origin as in the next sample:

Thus, little by little, in many cases, friendship bonds were bridged; they identified with each other since they were members of the same program and, in some cases, from the same geographical region. In spite of the strong identification among participants, some tensions were perceived due to personality clashes. For instance, while some tutees expected their tutors to be more open and share intimate issues with them, their guide may have kept reserved and limited her/his counseling to academic matters.

In addition to the previous considerations, tutees usually shared with tutors their concerns regarding various components of the program. They wanted to know about the curriculum, the university services, and the dynamics of certain courses. This interest generated constant reflection on the suitability of the program and their performance in their studies. Situations in which students considered staying in the program despite their difficulties were discussed in the sessions.

Tutees also expressed that this experience made them aware of the roles they could perform as future teachers or tutors since they analyzed how their tutors acted as guides. An issue to remark on at this point relates to tutees' expectations in regard to tutors' language knowledge and skills. They soon discovered that their guides might also have had some gaps in their competence to handle the language; however, their tutors' eagerness to become more prepared made pupils build a realistic image of what a teacher is. This implies they viewed their tutors as models to be followed. Their admiration for what the former ones did exemplified some of the qualities they could attain and the social commitment they would have to face when becoming teachers.

From Engaging to Acting: Shapes of Tutees' Willingness to Undertake Their Learning

Enrolling in the tutorial program was a free decision tutees took. As weeks passed, various situations revealed that their permanence in the program was tightly connected to their own inner drive to manage their learning. Little (1991), Low (1996), and Dickinson (1995) refer to this aspect as an evidence of autonomy. Thus, our interest in exploring how peer tutoring reveals pupils' autonomous work is the core of this section. As this kind of tutoring was a novelty for students, we shaped the main findings into a metaphor based on the process one follows in deciding which new suit one wants to wear.

Trying on a New Garment

Tutees' possibilities to exercise their autonomy revealed various issues. These facets had as their main pillar tutees' personality and learning factors. Initially, pupils' shyness, fear and lack of motivation restricted their autonomy.

When pupils were asked about the relationship between peer tutoring and autonomy, different expressions connected to motivation were mentioned, among them: "to be willing to do something, to have more interest in a subject, and desire to develop something one wants" (I Te 1-2 2007).

Becoming aware of what learning means, one's process and the possibilities to achieve a successful performance was another dimension of students trying to wear that new garment. Participants acknowledged the importance of attending tutorials as a step towards becoming agents of their own learning. Likewise, diagnosing their weaknesses and strengths, knowing about the development of tutorials and the means tutors suggested them to work with, were stages in tutees' journey towards autonomy.

Tailoring the Experience to Meet One's Needs and Objectives

As has been mentioned previously, one of the starting points in our tutoring cycle involved tutee's identification of their own learning objectives through their personal improvement plans. Data collected revealed that some pupils did not decide on their goals, but that these were set by tutors or born from immediate needs. On the contrary, a good number of tutees clearly expressed their aims to work on specific concerns from the beginning or progressively along their participation.

Tutees' decision-making in regard to their scope of intervention to reach their learning goals disclosed three progressive levels as they engaged in peer-tutoring. Figure 3 evidences our findings in relation to the possible connections between tutorials and pupils' autonomous development. The figure has been conceived as a pyramid which describes how the group of participants evolves in their decision-making process to reach their aims and illustrates each case by means of a piece of evidence. On its base, we have placed a group of tutees who shows an initial intention to explore the various possibilities they have at hand for learning. Pupils' involvement in tutoring seemed to support their progression towards a more active role in terms of selecting their learning opportunities when they moved to the second stage. Finally, tutees advanced towards developing a clear sense of what they needed and how to accomplish it on their own initiative. These pupils are located in the apex of the pyramid.

Along their participation in peer tutoring, first semester EFL student-teachers worked in an environment which provided them conditions to assume more positive attitudes towards their preparation. Pupils started to gain a more favorable view of their future roles as teachers; they regarded their guides as models to learn from. Not only were tutors' actions as language instructors analyzed by tutees, but they also paid attention to the values they bestowed thus tutees reflected upon important qualities for teaching. Likewise, tutees also gained information about their program and how they could function in it.

Additionally, tutees understood better how English works and how the nature of language bears implications in regard to its learning. Several misconceptions tutees held about language learning were initially reshaped due to their reflection on tutorial spaces. Peer tutoring, as a practice opportunity space, supported tutees' proficiency and their preparation since they transferred attitudes, skills, and knowledge from tutoring to the various classes taught in English.

Tensions were also experienced when tutees did not have a clear picture about the tutorial dynamics or the role they wanted to play in them. Conflicts arose around the kind of support they would receive, the logistics of tutorials and/or the relevance of using the target language. In some cases tutees' questioned their tutors' qualifications to guide them.

From proposing an agenda for reaching their objectives to evaluating their process, tutees always had the first chance to put their needs forward; that is why, as it will be inferred from the coming lines, several of the issues described up to now are factors which shaped tutees' autonomous learning. Progressing from a basic to a higher autonomous level, tutees moved along what can be compared to a pyramidal road. To begin with, some pupils depended excessively on their tutors and though they seemed to gain awareness about their process, most of their decisions in regard to their learning were made by their tutors.

A second level included tutees' gradual acceptance of their tutor's initial support which moved them ahead into expanding what they could do by themselves. The tutorial spaces exhibited examples or models with which they discovered different options to follow later on their own. These situations involved tutees' understanding of why and how they had decided to work in certain aspects or in a specific way.

Finally, we identified a third group of pupils characterized by their strong initiative to decide what to do and how to develop their tasks. They sought, elaborated and took materials of their preference to tutorials. They usually agreed with their tutors in having additional tutorial spaces.

Though the tutoring model has been shared and studied with tutors along various semesters, in some aspects there is still a gap between the theory behind it and the extent to which tutors follow its principles. In this sense, tutors can have a more solid component in their preparation to reflect upon crucial issues such as the attitudes which do or do not favor independence in tutees and the importance of being coherent with their role as a model. Being this a two-way road to build, we have also planned to include an initial meeting with tutees so they become more aware of the tutorial model principles and their commitment.

It is well-known, from theoretical sources in this area that peer tutors do not have to know it all since they are still consolidating their knowledge. Nevertheless, we are always looking forward to reinforcing their English language skills. That is why, in order to support tutors' language proficiency, we have organized English clubs, become ourselves their tutors or always had meetings in English. The constant request that tutees made for materials confirmed the need we have for more suitable resources to support their autonomous learning. Along the implementation of the tutoring model, we have worked in preparing many of the resources in our language laboratory to be used by tutors and tutees in our program (Viáfara & Ariza, 2008b).

The results of this study have determined that tutors had an impact in guiding tutees to consider their decisions about dropping out of the program. Bearing the previous finding in mind and that our university, as many others in our country, has looked for possibilities to retain students, we have planned to work more closely with specialized welfare departments in our university so that they can consolidate relevant orientation policies.

Since tracing in detail how tutorials impacted tutees' perfomance in classes emerged as a limitation in our study, structuring a peer tutoring model which bears a stronger link with program courses can be the focus for further study. This can initially imply analyzing in depth the connections between tutoring and language competence identified in this study, so that specific courses of action can be taken in the planning of English related subjects.

A fruitful area for further inquiry would be to determine how other variables interacted with tutoring in coaching students' exercise of their autonomy. The specific scope of our study, which sought to explore tutees' evidence of autonomous practices in tutoring, illuminated us exclusively to describe the tutees' autonomous profile and work in these spaces.

¹ I Te stands for Tutees' interviews.

² I Tu stands for Tutors' interviews.

³ TAS stands for transcription of an audio recorded session.

⁴ J stands for journal.