Lecciones a aprender desde el laboratorio en línea de enseñanza y aprendizaje de portugués como lengua extranjera

Cristina Martins (CELGA-ILTEC/Faculdade de Letras da Universidade de Coimbra)

Conceição Carapinha (CELGA-ILTEC/Faculdade de Letras da Universidade de Coimbra)

Celeste Vieira (CELGA-ILTEC/ Universidade de Coimbra ensino a distância da)

Artículo recibido: 13-10-2017 | Artículo aceptado: 14-11-2017

RESUMEN: En el contexto del proyecto E-LENGUA, la Universidad de Coimbra desarrolló un laboratorio online basado en la plataforma Moodle, con dos componentes: un curso online de portugués nivel A1 y un Corpus de Interacción Instructor-Aprendiz (IIA). Diseñado para promover el uso de las TIC en la enseñanza y el aprendizaje de portugués como lengua extranjera, el laboratorio ha impulsado el aprendizaje colaborativo de dos grupos de estudiantes con un dominio muy asimétrico del idioma. El objetivo del presente estudio es extraer las lecciones que se dedujeron de los conjuntos de datos recogidos: los niveles de participación en las actividades del curso online y las interacciones orales registradas en el ámbito del Corpus de IIA. Los resultados revelan la necesidad de medidas de mejora en el ámbito del diseño de instrucción y de la participación, sugiriendo fuertemente la idea de implementar actividades que promuevan la práctica reflexiva de los tutores online.
ABSTRACT: In the context of the E-LENGUA project, the University of Coimbra developed an online Moodle-based laboratory with two articulated components: a Portuguese as a Foreign Language (PFL) A1 level online course and an Instructor-Learner Interaction Corpus (ILI Corpus). Designed to promote the use of ICT in the teaching and learning (TL) of PFL, the PFL-TL Lab has stimulated the collaborative learning of two groups of students who exhibit very asymmetrical levels of language proficiency. The goal of the present study is to present the lessons that were drawn from two sets of data collected in PFL-TL Lab: learner participation levels in the online course activities and oral interactions recorded for the ILI Corpus. Results reveal the need for improvement measures targeted at instructional design and participants, strongly suggesting the need for the implementation of activities fostering the reflective practice of online tutors.

PALABRAS CLAVE: Enseñanza y aprendizaje del portugués como lengua extranjera, niveles de participación, interacción oral sincrónica, ruptura de la comunicación, asimetría de la competencia lingüística
KEY WORDS: Teaching and learning of Portuguese as a Foreign Language, participation levels, synchronous oral interaction, communicative breakdown, asymmetry of language proficiency

---

The E-LENGUA project is financed by the KA2013 Strategic Partnerships Actions for Higher Education.

---

1. Introduction

In the context of the E-LENGUA project, the University of Coimbra (UC) was faced with the challenge of presenting an action proposal aimed at fostering the effective use of computer-mediated asynchronous and synchronous communication systems in FL teaching to facilitate learning. This challenge was addressed through the development of an online Moodle-based laboratory, designed to promote the use of ICT in the teaching and learning (TL) of Portuguese as a Foreign Language (PFL). The E-LENGUA@UC action proposal has, in fact, originated two major intellectual outputs, a PFL A1 level online course and an Instructor-Learner Interaction Corpus (ILI Corpus), both of which are articulated components of the PFL-TL Lab. Data collected during online course delivery (a part of which are now included in the ILI Corpus), and throughout the very process of the PFL-TL Lab construction have provided valuable lessons for improving course design, instructor training and, ultimately, computer assisted language learning. In this study, these lessons and the data from which they have been drawn, namely participation levels in the online course activities and oral interaction analysis, are presented and discussed.

The PFL A1 level online course has provided an environment in which two groups of UC students can now meet and learn collaboratively: beginner PFL A1 level learners of the Erasmus mobility program, and MA students of the Portuguese as a Second and Foreign Language program. For both groups, the PFL-TL Lab is an optimal space for problem solving and learning by doing. MA students have faced the challenges of instructional design, while also gaining tutoring experience with genuine learners of PFL, and Erasmus students have been provided with opportunities for incidental learning of PFL, extra instructional time, and interaction sessions with native or near-native speakers of Portuguese who are likewise students, even if not exactly peers.

Given the fact that participation in the PFL-TL Lab activities is voluntary and free of charge for both groups of UC students, data regarding the levels of engagement and personal commitment during online course delivery are particularly relevant for assessing the effectiveness of instructional design. Such information provides valuable insight into the features of the instructional activities that are most likely to motivate learners who have willingly volunteered to spend some of their spare time taking an online course. Important lessons have thus been drawn from the participation levels registered for each course activity, and these will be discussed in section 4.1. of the present study. In this discussion, the impact of three main features will be considered: the linguistic skill involved in each activity, the technological tool used to implement it, and the degree of immediacy of the feedback provided to the learners during and/or after task completion.

Another very important source of information for assessing the effectiveness of the teaching and learning that takes place in the Lab is the ILI Corpus. The data included in the ILI Corpus (260 minutes of video-recordings) were collected during the synchronous oral interaction activities of the PFL A1 level online course, and they have revealed the extent to which the asymmetry in language proficiency between participants in oral interactions generates difficulties in communication that both parties must learn to manage as effectively as possible.

In real-life oral interactions, occurrences of communicative breakdown can be very costly, but in a protected and controlled environment, such as the PFL-TL Lab, they can represent valuable learning opportunities. Research has shown that, unlike one way input, interactional input provided by more proficient speakers to less proficient speakers in oral conversations is generally modified through a negotiation process, as to become more comprehensible to the less proficient speakers (Krashen, 1981; Long 1983; Gass et al., 1998). During this process, less proficient participants end up focusing more on form, as they are presented with alternatives for conveying the same meanings, thus highlighting function-form relationships in the target language. Furthermore, in an oral interaction, learners are compelled to produce output, hence facing the need to summon linguistic resources, experiment with lexicon and grammatical structures, and acknowledge linguistic and communicative difficulties that still need to be worked on (Swain, 2000). The Lab provides a safe learning environment where both PFL learners and tutors can negotiate, experiment and make mistakes without incurring major losses. Learners, depending on the nature of each course activity, are always provided with either immediate or delayed feedback. Tutors also benefit from both types of feedback, during the instructional design phase and the delivery of the course.

Of special interest to the present study, the analysis of the ILI Corpus has allowed for the identification of communicative breakdown signalling categories and repair strategies that are invaluable for sustaining tutor self-monitoring activities and the development of a best practices guide for novice instructors. Typologies resulting from this analysis are presented and discussed in section 4.2.

The experimental nature of the Lab is also manifest throughout its ongoing construction process. In effect, the 5 stage ADDIE Instructional Design method, used as a framework in designing and developing educational and training programs, was followed in the construction of the PFL-TL Lab. ADDIE stands for Analyze, Design, Develop, Implement, and Evaluate (Miranda, 2009):

1. Analyze – diagnose training/learning needs of target groups and identify general goals;
2. Design – formulate specific objectives and define the related activities;

• Develop – prepare contents and resources;

1. Implement – provide the learners with course materials and deliver the course;
2. Evaluate – assess all previous stages and reformulate anything that needs revision.

Presented as the fifth stage, evaluation takes place effectively throughout the entire ADDIE cycle, producing intermediate results between stages. Construction of the PFL-TL Lab has observed this principle, benefiting from ongoing evaluation and continuous feedback from all parties involved: PFL learners, tutors, and members of the supervising teams overseeing technical, scientific and pedagogical issues.

2. PFL-TL Lab: course characterization

During the 2015-2016 school year, the first module of the PFL-TL Lab, focusing on oral skills, was created, and the full cycle of the ADDIE model was completed over the course of 6 months.

In 2016-2017, a new group of tutors and PFL learners joined the Lab. Over the course of the academic year, new activities, primarily focusing on writing skills, were added to the existing course. Improvements suggested after the delivery of the first edition of the course were also implemented (especially regarding scoring criteria), and the second edition was delivered between March and April of 2017.

The current version of the course includes 17 activities (2 synchronous and 15 asynchronous), using different technology tools, and targeted at different language skills. The activities are distributed into 6 themes.

Theme	Activity	Communication: Asynchronous (A)/ Synchronous (S)	Language skill	Technical tools / Feedback
1.Personal data	1. Training pronunciation (optional)	A	Non-interactive oral production	Text to Speech Tool (SitePal) / Self-monitorization
	2. Videoconference	S	Oral interaction	Skype / Immediate
	3. Self-presentation video (optional)	A	Non-interactive oral production	File submission / Delayed
	4. Studying in a group	A	Non-interactive writing Written interaction	Forum (Q&A) / Delayed
2. Description of people, animals and objects	1. Descriptions of objects	A	Oral comprehension	Test / Immediate
	2. Who is who?	S	Oral interaction	Skype / Immediate
3. Family, houses and social relationships	1. Family tree	A	Oral comprehension	File submission / Delayed
	2. True or false	A	Oral comprehension	Test/ Immediate
	3. Family ties	A	Non-interactive writing	Test/ Immediate with subsequent tutor validation
	4. Rooms in a house	A	Non-interactive writing	Test/ Immediate with subsequent tutor validation
	5. João talks to his friends	A	Reading comprehension	Test / Immediate
4. Food	1. «Pastéis de nata» cooking recipe	A	Oral comprehension	Test / Immediate
5. Shopping	1. On sale	A	Oral comprehension	Test / Immediate with subsequent tutor validation
	2.Market dialogue	A	Oral comprehension Reading comprehension	Test / Immediate
6. Schedules and spatial orientation	1. What time is it?	A	Writing	File submission / Delayed
	2. The maze	A	Reading Comprehension	Test / Immediate
	3. A day in Coimbra	A	Reading Comprehension Non-interactive writing Non-interactive oral production	File / Delayed submission

Table 1: Characterization of the PFL A1 online course activities

Course activities cover a wide range of productive and receptive skills involving both oral and written forms of language: (i) oral interaction between tutors and learners; (ii) non-interactive oral production; (iii) oral comprehension; (iv) written interaction; (v) non-interactive writing; and (vi) reading comprehension. Different technological tools have been used to implement these activities, some of which are available in the LMS Moodle, such as Q&A forums, tests and file submission, while others are external to the Moodle, namely Skype and the text to speech SitePal tool.

Activities also differ depending on the type of feedback that is provided to learners. When feedback is delayed, it is always supplied by a human being (the tutor), but immediate feedback can be both automatic/machine generated or human. Activities implemented as open answer tests yield feedback of both kinds, in two stages, given the fact that the first automatic machine generated response requires validation by the tutor. For this reason, open answer tests fall under the delayed feedback category.

The combined effect of language skill, technological tool and type of feedback generates major groups of activities with the following clustered features: (i) writing + forums / file submission + delayed feedback; (ii) oral or reading comprehension activities + test + immediate automatic feedback; (iii) oral interaction + Skype + immediate human feedback.

3. Participants

The results discussed in the present study were extracted from data collected during the 2nd edition of the PFL A1 level online course in which 5 tutors and 23 PFL learners participated.

All tutors were female. Most (4) were originally from Portugal and their native language was Portuguese, and one was from Mozambique, with Changana as her native language and a near-native proficiency in Portuguese. Four were MA students of the Portuguese as a Foreign and Second Language program (ages between 20 and 40 years old), and one was an undergraduate student majoring in Modern Languages (under 20 years).

As for the 23 PFL learners, three were male and 20 were female. Their ages ranged from 20 to 30 years old, and they were originally from 10 different countries: France (1), Poland (10), United States (1), the Czech Republic (2), Israel (1), Slovakia (3), South Korea (1), Algeria (1), Hungary (2) and the Netherlands (1). The learners were all Erasmus students at UC, from different academic and scientific backgrounds, who, a month prior to their participation in the PFL-TL Lab, had just started a face to face A1 level course, delivered during 4 hours a week, in groups with a mean of 37.5 students.

4. Results

4.1. The PFL A1 online course activity participation reports

In the interest of improving instructional design, one very important source of data consists in the activity participation reports generated by the LMS Moodle.

Data plotted in chart 1 reveal that each learner participated, on average, in 12 out of 16 activities (activity 1.1 was not considered, given the absence of participation records in the LMS).

Participation levels tend to decrease during the life cycle of the online course, even if in a non-linear fashion. In fact, while it is true that stronger participation levels were recorded at the beginning of the course and that these become weaker towards the end, it is also noteworthy that strong participation in the early stages of course delivery is only manifest in activities that provide some type of immediate feedback (1.2, 2.1 and 2.2), and that the participation (P) / non-participation (NP) ratio is significantly inverted in asynchronous activities yielding delayed feedback, even at the beginning of the course (activities 1.3 and 1.4).

Chart 2 also illustrates a preference for activities based on oral skills (for which immediate automatic or human feedback is provided). Considering all features combined, it is furthermore clear that learners who voluntarily took the PFL A1 level online course preferred to spend their free time in synchronous oral interaction activities with immediate feedback from their tutors, a finding that is consistent with those of other studies: “Research data suggest that online courses with high levels of interactivity lead to higher levels of student motivation, improved learning outcomes, and satisfaction over less interactive learning environments” (Croxton, 2014: 316). Learning approaches that allow participants to ask and answer questions in the immediacy of real-time communication “provide opportunities for social interaction, peer/teacher support, and knowledge sharing not easily accomplished through asynchronous means” (Bower et al., 2015: 2).

Data analysis conducted thus far suggests the need for improvement measures focused both on participants and instructional design. One lesson learned from the PFL-TL Lab experience is that new activities to be added to the A1 level online course should rely as much as possible on immediate feedback. More demanding activities, involving writing skills, for instance, are better placed further ahead in the activities’ sequence, as not to present, too early on, a discouragement due to lack of adequate language proficiency.

Regarding learners, evidence also reveals the need for creating a stronger and more palpable academic incentive for participation in the course (through grading, for instance). As far as tutors are concerned, an autoscopy activity especially focused on skills for managing communication breakdown in synchronous oral interaction activities is highly recommended, as the results discussed in the following section clearly demonstrate.

4.2. The ILI Corpus

Given the asymmetrical level of proficiency in Portuguese of tutors (T) and learners (L), multiple occurrences of communication breakdown were to be expected. Understandably, learners signaled communication breakdown more often. Yet, and just as reported in other studies, the verbal performance of novice tutors is also a frequent source of miscommunication. Toyoda and Harrison (2002), for instance, analysed computer-mediated interaction between undergraduate learners of Japanese living in Australia and Japanese native speakers. The authors noticed that certain communication breakdowns were caused by tutors’ use of inadequate discourse features, such as abbreviated sentences, that are typical of face to face interaction, but representing a source of added difficulty during online chat communication.

How do learners signal a communicative problem they have encountered during the interaction? Table 2 presents the strategies found in our data.

Signals	Examples in Portuguese	Examples translated into English
Non-verbal: moments of silence, filled pauses, laughter	L – Eu gosto de…uhhh… T – … dançar, não é?	L – I like to….uhhh… T – …dance, right?
Metalinguistic comments	T – Vai estando atenta… L – Umm hum… T – …porque vamos falar por mensagem L – «Por mensagem»? What does it mean?	T – Pay attention… L – Umm hum… T – …because we’ll be communicating through message
Use of English or of codeswitching	L- I think it’s something about eating? We didn’t have this on my class so…I will not be able to answer…
	L – I don’t know what it is «Gosta»…

Table 2: Signalling breakdown

After the occurrence of the indicator (Varonis & Gass, 1985), this is to say the signal indicating that a communication problem has occurred, how does the tutor respond to the breakdown?

We should bear in mind that tutors are acting as instructors who must respond to the learners’ performance. At the same time, and in the specific case of oral interaction activities, they are also obliged to take charge of the dynamics of the interaction and are responsible for the smoothness of conversation flow. This double-faced role is uncomfortable for tutors (Zourou, 2009), and when breakdown does threaten the interaction, tutors take on full responsibility for preserving communicative continuity.

Following learners’ indicators, the tutors’ most widespread reaction is to continue speaking Portuguese. It must be noted that tutors were given minimal guidelines: they were provided beforehand with the grammatical and thematic syllabus of the face-to face class the learners were taking, and were instructed to speak Portuguese. So, although English or codeswitching were occasionally used as remedial tool (a common strategy otherwise reported in existing studies^[1]), in most cases, overcoming communication breakdown was attempted in Portuguese.

Overcoming communication breakdown can be attempted and achieved through positive remedial strategies. However, tutors may alternatively opt for negative management strategies that further aggravate the problem. When this was the case, tutors were found to adopt two different communicative behaviours: they either fled or fought.

In our data, fleeing is represented by changing the subject (table 3): the learner signals a problem and the tutor acknowledges it, but instead of attempting a corrective or clarifying move, he/she avoids the problem and suddenly changes the topic without introducing a framing or focusing move (Sinclair & Coulthard 1975: 22).

 

Strategy	Examples in Portuguese	Examples translated into English
Change the subject	T – Tu vieste para Portugal? L – No. T – Não? Então? Onde estás? Onde estás a conversar? Onde estás? L – *Bem. T – Onde estás? Estás em casa? L – *Na casa. T – Estudas português? L – Sim. T – Porquê? L – I don’t know. T – Gostas de línguas? L – No.	T – Did you come to Portugal?   T – No? So, where are you? Where are you talking? Where are you? L – *Well. T – Where are you? Are you at home? L – *In home. T – Are you studying Portuguese? L – Yes. T – Why?   T – Do you like (learning) languages?

Table 3: Tutor management strategy: change the subject

 

The introduction of framing or focusing moves helps the learner to expect a topic change, because they signal the transition between topics. On the contrary, the absence of these moves is deeply disturbing to the learner, given the abrupt changes of topic coupled with lack of topic closure. One of the assumptions that guide conversational/interactional behavior is that there must be some logical link between utterances. Grice (1975) referred to this shared expectation in his definition of the Maxim of Relation. When this maxim goes unobserved, the discourse is considered incoherent. As such, in these cases, the onus of the communicative problems encountered can also be ascribed to tutors. As stated by Boltoun and Booth (2001: 8), “all communication involves two active parties, and both necessarily have to share the responsibility for any breakdown.”

The alternative behavior exhibited by tutors consists in fighting to surpass the breakdown. In theory, fighting is positive, and it does often contribute to overcome the communitive problem at hand. However, other times fighting strategies result in mismanagement of breakdown.

In our data, tutors resort to different types of question sequences that are intended as remedial strategies in the face of communicative breakdown. Some of these sequences are successful in overriding the communicative disruption (see examples in table 4). A good example is the repetition of the previous question with rhythm variation and/or with auxiliary gestures. This is also the case of multiple choice questions, intended to simplify the learners’ task by providing him/her with possible alternative answers to the question that triggered the breakdown. In this sense, these supportive strategies can be considered scaffolding techniques provided by tutors to assist the learner. In accordance with the features that characterize scaffolded help (Wood, Bruner & Ross, 1976), these questions represent an attempt to simplify the task, they maintain the focus on the goal and demonstrate one possible or partial answer to the problem.

Strategy	Examples in Portuguese	Examples translated into English
Repetition a) identical b) slower c) with auxiliary gestures	b) and c): T – Queres dizer mais alguma coisa?’ L – (silence and smiles) Uhhh… T – Tu… queres… dizer (waves hand in front of mouth) mais alguma coisa?	T – Do you want to say anything else?   T – Do you… want… to say anything else?
Multiple choice questions	T – Como é que vais para a faculdade? (breakdown) T- Se vais, por exemplo, a pé… de autocarro…	T – How do you get to the faculty?   T – If you, for example, walk… take the bus…

Table 4: Successful management strategies: question sequences

However, other question sequences represent a mild to serious threat to conversational development (see examples in table 5).

Strategy	Example in Portuguese	Examples translated in English
Reformulation a) without structural simplification	T – Quer fazer-me alguma pergunta, Anna? L – (silence) T – Quer fazer-me alguma pergunta? L – (silence) T – Quer perguntar-me alguma coisa? L – (silence) T – A Anna tem uma pergunta para mim? (silence)	T – Do you want to ask me a question, Anna?   T – Do you want to ask me a question?   T – Do you want to ask me something?     T – Do you, Anna, have a question for me?
a) with structural simplification	T – Antes já tinhas vindo a Portugal? (breakdown) T – Já vieste a Portugal antes? (breakdown) T – É a primeira vez em Portugal?	T – Had you ever been in Portugal before?   T – Have you ever been in Portugal before? T – Is this your first time in Portugal?
False paraphrase: a) shift from WH to Y/N question	T – O que gostas de fazer nos tempos livres? Ficas em casa? Sais com os teus amigos?	T –What do like to do in your spare time? Do you stay at home? Do you go out with your friends?
b) maintenance of question type	Y/N questions: T – Gostas de estar em Coimbra? (breakdown) T – Gostas de Coimbra?   WH questions: T – Estás em Portugal para fazer o quê? (breakdown) T – Estás a estudar que curso?	T – Do you like being in Coimbra?   T – Do you like Coimbra?     T – What are you doing in Portugal?   T – What are you studying/majoring in?

Table 5: Unsuccessful management strategies: question sequences

Reformulation attempts, both with and without structural simplification, represent alternative forms of requesting the same information, thus providing enriched input to learners. This, in theory, is a good thing. Ideally, tutors’ efforts to clarify the intended meaning of the original question by presenting new sentence formulations aim to help the learner “to deliver more elaborate and substantial answers” (Heinonen & Lennartson-Hokkanen, 2015: 49). However, in reality, the juxtaposition of questions that follow the communicative breakdown does not keep with the tutor’s goals. Given the beginner status of A1 learners, reformulation is actually quite disruptive of information processing. Instead of helping the learner, this strategy puzzles him/her because the flood of questions represents too much information to process.

Even more so than reformulation, false paraphrase represents a very serious threat to conversational progression, for the sequence of questions, while falling under the same general theme, request, in effect, different items of information. Learners are obliged to shift their focus from one item to another, while they are also frequently required to process very different types of interrogative sentences in sequence. Adding to these difficulties in processing the input, output attempts are inhibited by the scarcity of linguistic resources available to the beginner learner who, presented with such demanding and chained requests, is unable to correspond with as little as one relevant answer.

5. Conclusions

The data collected in the PFL-TL Lab and examined in this study has contributed to a better understanding of the problems that may arise in computer mediated teaching and learning of languages. Part of the problems identified is related to instructional design and others are more directly concerned with participants’ shortcomings.

Activity participation data unveiled the relevance of favoring tasks involving more interactivity and providing immediate feedback. While these features may be present in tasks relying on automatic/machine generated feedback, our data has also made clear that learners prefer oral activities involving human interaction, yielding ongoing immediate feedback.

The same set of data has highlighted the need to revise activity sequencing and the interest in deferring the placement of more demanding activities, such as those requiring non-interactive writing skills, to later phases of course delivery, when, in principle, learners should be better equipped, as far as their linguistic proficiency is concerned, to engage more autonomously in these types of tasks.

Oral interaction data has further revealed the importance of tailored approaches to address the specificity of highly asymmetric proficiency levels of participants. While it is obvious that communication breakdown is an expected consequence of learners’ low proficiency, it is also true that tutors can benefit from pointers to improve their coping strategies when faced with communicative breakdown (Boulton & Booth, 2001). If we combine both causes of miscommunication (low proficiency level of learners and an over-demanding task imposed on MA students with minimum training), it can be argued that this study sheds light on key pedagogical features of task design and tutor training in an online learning context^[1].

The examination of question sequences used by tutors to override communicative problems reveals two categories that were particularly challenging for learners. Reformulation, both with and without structural simplification, seems like a good idea in principle, but it can, in fact, create additional stress for beginner learners during input processing. False paraphrase, in whatever form, is highly distracting, preventing learners from focusing during input processing and interfering with their capacity to summon the linguistic resources needed to answer even if only one of the cascading questions.

Results of our study suggest the greater effectiveness of remedial strategies that rely on the constancy of syntactic structure (through different forms of repetition) and on the availability of optional answers (provided in multiple choice questions). Learners can also possibly benefit from pre-tasks focusing on the lexicon and the grammatical structures most likely to be used during the interaction. Data further reveals the importance of introducing framing and focusing moves as to structure the conversation and to provide learners with robust and unequivocal cues for topic change.

Tutors can benefit greatly from learning explicitly about what can go wrong during oral interactions with low proficiency learners and about what can be done to override the difficulties that will most certainly be encountered. The development of a reflective practice, this is to say knowledge construction through action and reflection before, during and after action, is the goal of the autoscopy activity to be implemented in the sequence of these findings, in the hope that this exercise will contribute effectively to the reinforcement of the novice instructors’ autonomy.

6. Works Cited

Boulton, Alex and Booth, Paul (2001). “Communication in the void and communication avoided: a case study of on-line language teaching”. ASp <http://asp.revues.org/1877> [10-10-2017].

Bower, Matt, Dalgarno, Barney, Kennedy, Gregor E., Lee, Mark J.W. & Kenney, Jacqueline (2015). “Design and implementation factors in blended synchronous learning environments: Outcomes from a cross-case analysis”. Computers & Education 76: pp. 1-17.

Croxton, Rebecca (2014). “The Role of Interactivity in Student Satisfaction and Persistence”. MERLOT – Journal of Online Learning and Teaching 10 (2): pp. 314-325.

Gass, Susan, Mackey, Alison & Pica, Teresa (1998). “The role of input and interaction in second language acquisition: introduction to the special issue.” Modern Language Journal, 82 (3): pp. 299-307.

Grice, Herbert Paul (1975). “Logic and conversation”. In: P. Cole and J.L. Morgan (eds.). Syntax and Semantics 3: Speech Acts. New York: Academic Press. pp. 41-58.

Heinonen, Maria Eklund & Lennartson-Hokkanen, Ingrid (2015). “Scaffolding Strategies: Enhancing L2 Students’ Participation in Discussions about Academic Texts”. Journal of Academic Writing 5 (1): pp. 42-51.

Kötter, Markus (2003). “Negotiation of meaning and codeswitching in online tandems”. Language Learning & Technology 7 (2): pp. 145-172.

Krashen, Stephen (1981). Second Language Acquisition and Second Language Learning. Oxford: Pergamon.

Long, Michael (1983). “Linguistic and conversational adjustments to nonnative speakers”. Studies in Second Language Acquisition 5: pp. 177–194.

Miranda, Guilhermina (2009). “Concepção de Conteúdos e Cursos Online”. In: Ensino Online e Aprendizagem Multimédia. Lisboa: Relógio D’Água Editores.

Sinclair, John, & Coulthard, Malcolm (1975). Toward an Analysis of Discourse: The English Used by Teachers and Pupils. London: Oxford University Press.

Swain, Merrill (2000). “The Output Hypothesis and beyond: Mediating Acquisition through Collaborative Dialogue”. In J. P. Lantolf (ed.). Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press. pp. 97-114.

Toyoda, Etsuko and Harrison, Richard (2002). “Categorization of text chat communication between learners and native speakers of Japanese”. Language Learning & Technology 6 (1): pp. 82-99.

Varonis, Evangeline Marlos & Gass, Susan (1985). “Non-native / non-native conversations: A model for negotiation of meaning”. Applied Linguistics 6: pp. 71-90.

Wood, David, Bruner, Jerome S. & Ross, Gail (1976). “The role of tutoring in problem solving”. Journal of Child Psychology and Psychiatr 17: pp. 89-100.

Zourou, Katerina (2009). “Corrective feedback in telecollaborative L2 learning settings: Reflections on symmetry and interaction”. The jalt cal Journal 5 (1): pp. 3-20.

· Descargar el vol.6 nº2 de Caracteres como PDF. · Descargar este texto como PDF. · Regresar al índice de la edición web.