System 62 (2016) 63e76

Contents lists available at ScienceDirect

System

journal homepage: www.elsevier .com/locate/system

Benefits of mobile instant messaging to develop ESL writing

Alberto AndujarDepartment of Linguistics, University of Almería, Spain

a r t i c l e i n f o

Article history:Received 2 June 2015Received in revised form 8 July 2016Accepted 11 July 2016Available online 22 July 2016

Keywords:Mobile instant messagingChat-based conversationSecond language writingMobile assisted language learningNoticingCollaborative learning

E-mail address: aav447@ual.es.

http://dx.doi.org/10.1016/j.system.2016.07.0040346-251X/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

This study investigated the benefits of Mobile Instant Messaging (MIM) through ananalysis of grammatical, lexical and mechanical accuracy as well as syntactic complexity insecond-language learners' writing. A WhatsApp group was created where 80 Spanishstudents taking a B1 English course participated in a daily interaction during six months. Aquasi-experimental research design with an experimental and control group and a pre-post test was followed. Students were divided into two main groups according to treat-ment type with 40 students in each group. This research focused on the interaction in theapplication and attempted to measure, through a qualitative and quantitative analysis, thestudents' degree of writing development. The ratios of lexical, grammatical and mechan-ical errors as well as error-free clauses per clause and error-free T-unit per T-unit indicatedsignificant differences between the control and experimental group in terms of accuracy.Nevertheless, measures of syntactic complexity together with lexical diversity were notconclusive as the independent parameters for syntactic complexity showed no significantdifferences between the two groups. WhatsApp constitutes a powerful educational tool toencourage second language interaction among participants and its tremendous potentialto activate students' involvement remains one of the least exploited functionalities ofmobile phones.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Mobile learning breaks temporal and spatial lines, placing educative resources at the service of the students. Throughmobile phones, Internet users now have the possibility of accessing the Internet from different points, having access to the netfrom almost any place. However, although its importance as a research field is increasing rapidly, as reflected in the numerousresearch articles, conferences, seminars and workshops taking place all over the world (see Gülbahar, Jacobs, & K€onig, 2015),several reviews (see Burston, 2014; Cobcroft, Towers, Smith, & Bruns, 2006; Trifonova, 2003) indicate that MALL remainsmarginal, as experiments and class trials carried out lack follow-up reports showing curricular integration.

This study is based mainly on the need for extending research related to the use of Mobile Instant Messaging (MIM) in theEnglish as a Second Language (ESL) classroom and its benefits. MIM is an asynchronous and synchronous communication toolthat uses wireless networks and handheld devices, allowing students to hold conversations and exchanges (Dourando, Parker,& De la Harpe, 2007; Rambe & Bere, 2013). This type of messaging is characterized by instant delivery through a pop-upmechanism, a list of users, and a mechanism to indicate when they are available. At the same time, it allows access todiverse content, acquisition of specific data and information sharing among users. MIM constitutes one of the functionalitieswhich holds great potential in a mobile phone; typed messages have evolved to become a new hybrid of spoken, written and

A. Andujar / System 62 (2016) 63e7664

electronic chat discourse (Blake, 2000; Muniandy, 2002; Warschauer, 1996; Werry, 1996). Through this technology, studentsfind an environment where their interactions facilitate acquisition of second language communication skills (Blake, 2000;Bonk & Kim, 1998).

Despite its tremendous potential to activate students' involvement, MIM remains one of the least exploited functionalitiesof mobile phones. This could be explained by several reasons such as (1) the difficulty and cumbersomeness of typing on asmall mobile screen, as well as the screen resolution itself (Alexander, 2004; Shudong & Higgins, 2006), (2) the distractivenature of text messages, (3) limited academic conceptualization of how text messages can be integrated in ordinary in-struction, and (4) lack of certainty that the comments generated via text message have real academic rigour (Rambe & Bere,2013). Regardless of these hypotheses, this social practice allows contact with new information, builds social networks,supports the sharing of ideas and promotes mutual understanding and learning in informal contexts. Due to the inherentcharacteristics of mobile devices, this practice allows synchronous and asynchronous interaction virtually at any time andplace.

Focusing on these interactions, this study attempted to measure, through a qualitative and quantitative analysis, thestudents' degree of writing development. In particular, this research explored the pedagogical value of WhatsApp, its assetsand its academic potential.

2. Literature review

In order to contextualize this research, it is necessary to take into consideration notions such as communicativecompetence, as well as which communication skills this study focuses on. Subsequently, we will explain the theoreticalmodels that underlie the research.

From a functionalist perspective, communicative competence is defined as ‘an instrument of communication whoseproperties can be studied as long as they are instruments for communication’, as stated by Hymes (1972). Other authors likeCanale (1983) and Canale and Swain (1980), point out that communicative competence is the combination of knowledge andskills (linguistic, sociocultural, discursive and strategic) which allow an effective performance of the speaker, suitable for aparticular communicative situation and a specific speaking community.

Following the integrated approach espoused by Oxford (2001), four primary skills used to reach communicativecompetence can be distinguished in the EFL/ESL classroom: reading, writing, listening and speaking. We will take intoconsideration those belonging to the written medium (writing and reading), paying special attention to the writing skill.

2.1. Framework and Vygotsky's zone of proximal development

The Framework for the Rational Analysis of Mobile Education (FRAME) (Koole, 2009) describes ‘mobile learning’ as aprocess resulting frommobile technologies' convergencewith human capacities of learning and social interaction. This modeltakes into consideration the technical characteristics of mobile devices as well as social and personal aspects of learning(Koole, 2009). The theory underlying the Framework echoes Vygotsky's (1978) theorization about mediation and zone ofproximal development. Vygotsky's social development theory stresses the fundamental role of social interaction in thedevelopment of cognition (Vygotsky, 1978), as he deeply believed that community plays a central role in the process of‘making meaning.’ Therefore, in this sociocultural approach to cognitive development, individual development cannot beunderstood without reference to the social and cultural context withinwhich it is embedded (Lantolf& Thorne, 2006). Hence,higher mental processes taking place in the individual have their origin in social processes. This theory directly affects ourstudy, as the students are part of a social activity in which interaction plays a key role.

FRAME (Koole, 2009) underlines the role of technology and, at the same time, gives an important emphasis toconstructivism. In this model of learning, students can use different physical and virtual situations, interacting with otherpeople, information or systems at any time and place. Learning experiences take place in an information context whereinteraction is mediated through technology.

2.2. Mobile text chat and interaction

Swain (1985) regarded interaction as a key element in second language development since interaction constitutes afundamental factor where students construct knowledge in an active manner. This constructivist approach is based onbuilding active meaning, focusing the process on primary learning concepts and not on isolated facts.

Due to their real-time nature, text chats and consequently MIM resemble face-to-face interaction, and thus may carrylanguage development benefits such as repair moves or negotiation for meaning (Smith, 2003; Sotillo, 2000; Warschauer,1996). Repair moves are more frequent due to the fact that participants can send messages simultaneously regarding un-related or abandoned topics, hence creating a discourse sequence different from face-to-face interaction where they speak inturn on a single topic (Doughty & Long, 2003; Negretti, 1999). These messages appear within seconds on their interlocutors'screens and each textual turn contains the learner's language and typographical errors (Jepson, 2005). The negotiation ofmeaning that is produced when students suffer or perceive problems with communication is defined as a ‘process in which alistener requires message clarification and confirmation and a speaker follows these requests, often through repeating,

A. Andujar / System 62 (2016) 63e76 65

elaborating or simplifying the original message’ (Pica, 1994, p.497). Their discourse is modified in order to reach an under-standing, turning into comprehensible input that will rise to intake acquired by the students.

Research on task-based computer-mediated text chats has shown that about one third of the total turns taken by Englishlearners are related to negotiation, and it engenders SLA among NNs (Non-native speakers) (Chen, Belkada,&Okamoto, 2004;Pellettieri, 1996, 2000; Smith, 2003, 2004). In studies such as Payne and Whitney (2002) about task-based chats, Spanish L2participants of an experimental teacher-led group demonstrated greater improvement in their oral proficiency than those inthe control group. Furthermore, as several studies show, text chats seem to allow participants from different levels to help oneanother in co-constructing social activity (Ortega, 1997; St. John& Cash, 1995; Warner, 2004) as well as knowledge, followingthe cooperative learning theory stated by Slavin (1983). Conversely, although these studies address task-based teacher-ledchats whichmay seem similar toMIM, some distinctive features of MIM can be addressed: (1) synchronous and asynchronousinteraction, as opposed to CMC where only asynchronous interaction is carried out, (2) easy access to content and possibilityof seeing online users, (3) wider range of chat features such as negotiation of meaning, repair moves, comprehensions andconfirmation checks, clarification requests and self-repetition, (4) greater ease in media sharing: voice, images and text, and(5) availability to use Internet resources.

Furthermore, students do not use technology they consider intrusive or if they need time to learn how to use new toolsand software (Kukulska-Hulme & Shield, 2007). In contrast, WhatsApp is an application that many students have alreadymastered, which is a key advantage to avoid these drawbacks previously mentioned. Some of the assets of this applicationare: promoting contact between students and teachers; fostering interactions among students and promoting academiccooperation; encouraging active learning; providing instant feedback; developing high communicative expectations (Desai&Graves, 2006; Dourando et al., 2007; Farmer, 2003; Rambe & Bere, 2013).

Research in the field of MIM is very scarce. Nevertheless, some studies, such as Levy and Kennedy (2005), have exploredthe possibilities of SMSmessages in mobile devices. In their study the instructor sent messages at specific times and days, nottaking into consideration that students could answer anytime and anywhere. Furthermore, V�azquez-Cano et al. (2015) foundthat in MIM the amount of shortened words was reduced, as opposed to SMS where students did not have a word limit. Theirresearch investigated the linguistic and paralinguistic characteristics of MIM in Spanish teenagers and emphasized a su-prasegmental enrichment due to the use of emoticons and stickers. Other studies such as Diaz (2014), Morat�o Pay�a (2014) orPadron (2013) focused on the potential that WhatsApp offers to promote collaborative learning but failed to demonstratelinguistic gain.

2.3. Noticing and collaborative learning

Regarding the kind of learning that is produced through MIM, DeKeyser and Larson-Hall (2005) describe implicit learningas that which is produced without any awareness that such learning is taking place. This type of learning is characterized asbeing unintentional, unconscious and, on some occasions, involving consciousness, but only at the level of perception. Thisperception is considered by Schmidt (1995) to be necessary for the transformation of input into intake. Noticing is amechanism that mediates between communication and acquisition (Gass, 2003), where students notice the gap betweentheir current language and their target language, appreciating the feedback given by the teacher. In an output-input cycle,learners are pushed in their production and can compare it with the input subsequently received from other students as wellas the teacher in the chat group. This comparison aids learners in seeking relevant linguistic forms in order to conveymeaningmore accurately (Swain & Lapkin, 1995; Swain, 1998; Thornbury, 1997). According to Hanaoka and Izumi (2012), noticingengages learners' cognitive processing and they become aware of the possibilities as well as problems in their IL (Inter-language) capabilities and development, which will alter the subsequent input processing. Throughout the interaction,students explicitly focused their attention on language, questioning their language use or correcting themselves or others,which Swain and Lapkin (1998) called language-related episodes (LREs). In LREs attention is drawn towards formal features oflanguage, which leads to L2 acquisition (Izumi & Bigelow, 2000; Izumi, 2002; Izumi, Bigelow, Fujiwara, & Fearnow, 1999).Hypotheses are created by the learners who experience a cognitive process of making meaning, and shaping knowledge andexperience through reflecting on language, a process called ‘languaging’ (Swain, 2006), which becomes a source of L2learning. According to Dobao (2012), novice learners construct knowledge in collaboration with more capable individuals, inthis case, the teacher. Other studies have pointed out the benefits of learner-learner interactionwhere participants act as bothnovice and expert (Ohta, 2000, 2001; Storch, 2002; Swain & Lapkin, 1998).

2.4. Research questions

This study seeks to answer the following questions: (1) has the programme implementation in the experimental groupinvolved important effects in the development of accuracy as well as lexical and syntactic complexity of the subjects? (2) Isthe use of Mobile Instant Messaging profitable for this development? Regarding this last question, the study tries to answeraspects in relation to the potential value of the application in ESL writing as well as how this MIM provides opportunities forinteraction and, consequently, second language development.

A. Andujar / System 62 (2016) 63e7666

3. Methodology

3.1. Research design

The purpose of this exploratory research was to investigate writing development in ESL learners, through analysing theinteractions in a mobile text-based teacher-led chat. Regarding the research design, a quasi-experimental design with anexperimental and control group and a pre-post test was followed (Cambell & Stanley, 1963). The experimental group un-derwent the prescribed treatment while the control group received no treatment at all and served as the benchmarking pointof comparison. Naturalistic observation was used as the researcher looked at the behaviour and interaction of students in anatural setting with no attempts at intervention (Nunan, 1992).

Null and alternative hypothesis were posed and tested in order to reveal significant differences between two pre-existinggroups (control and experimental group).

� Alternative hypotheses: The level of accuracy as well as lexical diversity together with syntactic complexity of the studentsin the treatment group will be higher than those in the control group.

� Null hypothesis: No significant differences will be found between in terms of accuracy, lexical diversity and syntacticcomplexity in the two groups.

Participant selection and initial differences were determined by the tests given at the beginning of the course to bothgroups. Internal threats to validity (Campbell& Stanley, 1963) and effects such as the Hawthorne effect (Mayo,1933), the haloeffect (Thorndike, 1911) and participant expectancy were minimized as the students were not conscious that they were beingobserved and the teacher did not participate in the correction of the tests given. An external raterwas used in this process. Thetwo groups did not have the same timetable, the experimental one met in the mornings and the control in the evenings, thusstudents did not know the different activities carried out in both groups as they did not know each other.

3.2. Setting and participants

This study involved four groups of third year students from the University of Almeria that were taking a B1 English course.The materials and contents provided in the course met the parameters established by the Common European Framework ofReference for Languages (CEFR) for a B1 level. The expected entry level was lower than B1 as students who enrolled on thecourse aimed to achieve such level by the end of the course. Enrolment was determined by a A2þ placement test, which wasadministered to the students at the beginning of the course in order to ensure theywere able to take it. In case participants didnot pass the test, they were not allowed to enrol on the course. The length of this course was six months, from November toMay, and met for four and a half hours per week. The number of students was 80, 61 female and 19 male students, all of themwith the same nationality and ages ranging from 20 to 26. Students were divided into two main groups according totreatment type with 40 students each. Intact classes were arbitrarily assigned as experimental and control groups. Theresearcher participated actively in the process, as hewas the teacher of the groups studied. Although the control group did notreceive any treatment, it was used for studying the differences between groups after the implementation. Pre-testing wasundertaken to establish any differences between the control and experimental group, discussed below. In this manner, wewere able to measure the effect of the treatment on experimental group subjects. After applying this treatment or imple-mentation in the programme of the experimental group, we verified its effect through the comparison of the results obtained.The identities of the participants have been changed so as to maintain confidentiality in this research, meeting ethicalguidelines set by the American Psychological Association (2002).

In both groups, during formal learning in class, the amount of opportunities to speak, as well as to be in contact with otherstudents was very limited due to 1) lack of time, 2) psychological distance between teacher and student, and 3) students' lackof confidence to ask questions in public.

3.3. Study

This study focuses on determining the potential of MIM services like WhatsApp in order to develop ESL writing. TheWhatsApp application is presented as a tool to encourage participation and interaction during the course. WhatsApp is aninstant messaging platform that uses the Internet to connect different mobile phone users, allowing the creation of groups inwhich different users can interact. Exchanges between users can be produced in a synchronous or asynchronous manner, andare received on phones as alerts. In this research,WhatsApp did not substitute for explanations in class, but was used as a toolto maintain communication out of class on a daily basis, becoming a constant support for language use. All the students in theexperimental group participated in the activity, as they all had mobile phones with a 3G Internet connection and an availablewi-fi at the University as well as the WhatsApp application previously downloaded. The expenses of 3G Internet connectionwere covered by the participants whichmay influence the accessibility of such affordances. Notwithstanding that it could be amajor drawback, students made use of their own data plans when necessary as they were already using them before thebeginning of the activity and in any case did not consider participation in the activity to be costly. After creating theWhatsApp

A. Andujar / System 62 (2016) 63e76 67

group, the activity was carried out taking into account a series of conditions established by the teacher: participation in theactivity and English use was compulsory; a different question had to be formulated by a different student each day, 7 days perweek; the topic of the question was free; each student had to give at least one answer per question and there was nominimum or maximum text length; sharing images and voice messages in English was allowed; the teacher participated as astudent, responding to the questions. Availability of the teacher in the application was complete, and errors were not cor-rected in an explicit manner during interaction.

Regarding participation, a daily tracking of the minimum answer per student required was carried out in order to ensurestudent's involvement in the activity. Participants sent 13,937messages during the activity, which shows a remarkable degreeof engagement. Grading was not determined by participation, hence no sanctions were imposed if they did not participate asthe teacher tried to keep the activity as natural and authentic as possible in order to ensure reliability. Students thought theywere just chatting but not being observed or evaluated.

3.4. Experimental methods

Due to the short-length of the samples in the application, a chronological axis was used tomeasure the difference betweenboth groups. Written English tests at the beginning and end of the course were used to measure writing skill in a quantitativemanner, by analysing grammatical, lexical and mechanical accuracy as well as syntactic complexity together with lexicaldiversity. The model used attempted to measure B1 writing in which students can write simple texts such as letters, events,experiences or impressions on topics which are familiar or of personal interest by linking a series of shorter discrete elementsinto a linear sequence as stated by the CEFR. This model consisted of: two texts to write which were the same for both groupswith a certain number of words per text (100 words). Compositions were done in a classroom environment with a time limitof 30 min. Students of both the control and experimental groups were given these tests whose content dealt with a previoustopic explained during the course. The purpose of these tests was to become a tool to measure the students' writing level.Such tests proved to be very helpful, not only by allowing a comparison between the experimental and control groups, butalso by measuring the development of experimental group students in relation to their tests at the beginning and end of thecourse.

Following previous research of written texts (Dobao, 2012; Storch & Wigglesworth, 2007; Wigglesworth & Storch, 2009),accuracy, lexical diversity and syntactic complexity were analysed. Fluencywas not taken into consideration, as the number ofwords in each test was set in advance (100 words). Different measures of syntactic complexity were used: number of wordsper clause and per T-unit; number of clauses per T-unit produced by the students. These units of measurement analyze howstudents increase the syntactic complexity of the texts written, turning short discrete elements into a linear sequence asrequired by the CEFR for a B1 level. Furthermore, thanks to the increase of such complexity, students are able to meet B1 ‘can-do’ descriptors such as the possibility of writing essays, opinions, brief reports or stories as well as detailed descriptions,feelings and experiences.

Regarding lexical diversity together with syntactic complexity (see Skehan, 2009), a mean segmental type-token ratio(MSTTR) was calculated. The written texts were divided into segments of 50 words each, which were used to calculate theMSTTR of each student (see Dobao, 2012).

With regard to linguistic accuracy, three different types of errors were calculated: grammatical errors; lexical errors; andmechanical errors.

Grammatical errors. Participants' mistakes regarding grammar such as choice of verb tense, use of infinitives, subject-verb agreement or use of the auxiliary verb are considered. Words in bold indicate errors made by students.

Example 1

Maria: I will to travel to Madrid next weekend.Ana: Madrid is my favourite place. I usually go when I was younger.Maria: It's fantastic! Best place ever!

Lexical errors. Mistakes made due to the lexical choice of the participants are taken into consideration.Example 2

Nuria: Are you similar to your parents?Noelia: Yes! To my father … he is very active and easy-going but my mother is very bored.Nuria: Really? Why is she boring?Noelia: Because she is all the time sad and crying. She is very sensible.

Mechanical errors. Learners' spelling and punctuation problems are analysed throughout the interaction.Example 3

Sara: Is it dificult to learn English?David: English is not difficult for me Sara. However (,) pronunciation sometimes is very different.

TabPre

EEEEEE

a

b

A. Andujar / System 62 (2016) 63e7668

Sara: Yes for me to. I think Spanish is easier.The ratios of error-free clauses to total clauses, error-free T-units to total T-units and errors towordswere also calculated tomeasure accuracy.

3.5. Qualitative methods

Through the language samples obtained in WhatsApp use, we can observe the individual development of each student,thanks to the analysis and tracking of the LREs taking place in the application. These LREs arose as a result of the interactionwhat encouraged learners to collaborate in order to reach correct solutions (Swain & Watanabe, 2012). Conversely, thisprocess may have ended up generating non-target like forms, hence, feedback was required in order to guide the learningprocess (Hanaoka & Izumi, 2012). Students participated by helping partners with difficulties when using the language,reflecting on the language use and constructing knowledge. Mistakes were corrected by students throughout the interactionas well as by the teacher who monitored the activity in the application. This correction was implicit as both students andteacher rephrased the mistakes made by other classmates. Students' performance was not marked in the application due theshort length of text messages, which made it impossible to evaluate the conversations consistently. Students' answers, ifnecessary, were reformulated, and following Hanaoka and Izumi's (2012) recommendation for lower proficiency learners,reformulated parts were emphasized helping participants' noticing. Reformulation by the teacher was used throughout theinteraction providing instant feedback to the participants and promoting learners' spontaneous noticing (Cohen, 1983). Suchreformulation addressed different textse short mobile text messages - from those reformulations carried out in other studies(e.g., Adams, 2003; Lapkin& Swain, 2004; Lapkin, Swain,& Smith, 2002; Yang& Zhang, 2010) where the aimwas proficiency.Nevertheless, such reformulation of text messagesmay lead to changes in learners at all levels: lexical, syntactic, grammatical,clarity of thought, etc.

4. Results and analysis

4.1. Results

The ANOVA related samples repeated measures test was conducted to confirm that the differences between pre and post-test scores were statistically significant. Alpha level of significance was set at 0.05 under a nondirectional (two-tailed) hy-pothesis. As shown in the pre-test results, no significant initial differences were found between groups. The improvement inaccuracy scores by experimental and control group students are presented in Table 1. Significant differences were foundbetween the experimental and control group in pre and post-test regarding error-free clauses per clause (F (1,78) ¼ 8.766;p < 0.05) and error-free T-units per T-unit (F (1,78) ¼ 6.485; p < 0.05). Sidak adjustment was used to counteract multiplecomparisons, showing evidence of a greater improvement in experimental group students' accuracy in the post-test: (0.19;p < 0.05) error-free clauses per clause and ¼ þ 0.17; p < 0.05) error-free T-units per T-unit.

As Table 2 demonstrates, statistically significant differences were found between experimental and control group in theindependent analysis of grammatical, lexical and mechanical errors. The ratios of grammatical errors per word (F(1,78) ¼ 41.194; p < 0.05), lexical errors per word (F (1,78) ¼ 115.508; p < 0.05), and mechanical errors per word (F(1,78) ¼ 115.50; p < 0.05) are significant between groups in pre and post-test. Sidak was used to compare pre-post measuresin both groups, showing a significant decrease in the number of grammatical errors per word (�0.021; p < 0.05), lexical errorsper word (�0.032; p < 0.05), and mechanical errors per word (¼ �0.032; p < 0.05) of the experimental group.

The results of syntactic and lexical complexity analysis are presented in Table 3. No statistically significant differences werefound in the total number of T-units between groups. Nevertheless significant differences were found in the number of T-units between the first measure (pre-test) and the second (post-test) which show that participants' number of T-unitsdecreased during the course (F (1,78) ¼ 13.670; p < 0.05). An explanation for this drop in the number of T-units could be thehigh number of single clauses that participants used in the first measure, whichwas diminished in the post-test. As previously

le 1-post test measures of accuracy for texts written in control and experimental group.

Control group Experimental group Control groupa Experimental groupa

Total Mean SD Total Mean SD Total Mean SD Total Mean SD

rror-free clauses 194 4.85 2.68 206 5.15 3.5 278 6.95 2.29 341 8.53 2.65rror-free clauses per clause 0.28 0.19 0.31 0.22 0.43 0.15 0.61b 0.2rror-free T-units 140 3.5 2.3 159 3.97 3.03 201 5.03 1.9 283 7.08 2.2rror-free T-units per T-unit 0.32 0.2 0.33 0.25 0.49 0.19 0.68b 0.16rrors 646 16.15 3.7 608 15.2 2.83 449 11.23 3.1 236 5.9 2.02rrors per word 0.16 0.37 0.15 0.28 0.11 0.31 0.06b 0.2

Post-test results in experimental and control group.Statistically significant difference between experimental and control group (p < .05).

Table 2Pre-post test measures of grammatical, lexical, and mechanical accuracy for texts written in control and experimental group.

Control group Experimental group Control groupa Experimental groupa

Total Mean SD Total Mean SD Total Mean SD Total Mean SD

Grammar errors 231 5.77 1.86 212 5.3 1.65 161 4.02 1.31 76 1.9 0.81Grammar errors per word 0.06 0.02 0.05 0.02 0.04 0.01 0.02b 0.01Lexical errors 232 5.8 1.72 223 5.58 1.79 169 4.23 1.67 40 1.1 0.64Lexical errors per Word 0.06 0.02 0.06 0.02 0.04 0.02 0.01b 0.01Mechanical errors 185 4.62 1.56 176 4.4 1.05 118 2.95 1.1 79 1.98 0.92Mechanical errors per word 0.05 0.02 0.04 0.01 0.03 0.1 0.02b 0.01

a Post-test results in experimental and control group.b Statistically significant difference between experimental and control group (p < .05).

Table 3Pre-post test measures of complexity for text written in control and experimental group.

Control group Experimental group Control groupa Experimental groupa

Total Mean SD Total Mean SD Total Mean SD Total Mean SD

Clauses 682 17.05 3.08 672 16.8 3.96 646 16.15 2.68 562 14.05b 2.71T-units 438 10.95 2.85 476 11.9 3.44 413 10.33 2.7 416 10.4 2.96Words per clause 5.86 0.67 5.95 0.58 6.19 0.6 7.11 0.54Words per T-unit 9.13 1.97 8.4 1.89 9.68 1.87 9.61 1.8Clauses per T-unit 1.55 0.31 1.41 0.27 1.56 0.28 1.36 0.22MSTTR 0.68 0.07 0.68 0.05 0.75 0.05 0.82b 0.04

a Post-test results in experimental and control group.b Statistically significant difference between experimental and control group (p < .05).

A. Andujar / System 62 (2016) 63e76 69

mentioned, participants reduced the number of T-Units in both groups, however the quantity of words per T-unit in bothexperimental (1.21) and control (0.55) groups increased, although none of these differences were statistically significant.

Concerning the total number of clauses, significant differences were found between groups in the pre and post-test (F(1,78) ¼ 30.147; p < 0.05). Sidak test for multiple comparisons showed a significant decrease in the number of clauses in theexperimental group (¼� 2.1; p < 0.5). However, the number of words per clause increasedmore in the experimental group (¼1.16) than in the control (¼ 0.24), suggesting greater complexification, but yielded no statistically significant difference.

Lastly, a mean segmental type token ratio (MSTTR) was calculated and significant differences were found between groupsin pre and post-test (F (1,78) ¼ 10.038; p < 0.05). Although both groups showed greater ratios in the MSTTR, Sidak showed asignificant increase in the experimental group (¼ 0,64; p < 0.05), therefore more lexical diversity together with syntacticcomplexity in the treatment group is regarded.

These analytic procedures and the results obtained from them illustratemajor improvements in the number of errors at alllevels (grammatical, lexical and mechanical), with the experimental group being significantly more accurate than the control.Furthermore, syntactic complexity and lexical diversity are observed in the experimental group, as the figures in the MSTTRshow. Nevertheless, othermeasures of complexity such as words per clause andwords per T-unit, which showed better scoresin the experimental group, were not statistically significant.

4.2. Qualitative analysis

In order to show the evolution of the experimental group students, we made use of the conversations kept throughWhatsApp, particularly several samples of LREs taking place in the application. These LREs were tracked in order to see ifstudents reflected on the focus of each LRE in latermessages, correcting the errors previouslymade. Three random samples ofLREs were chosen and later messages of the students involved in such LREs were observed.

Fig. 1 shows a LRE in which the student ‘Marta Alo B1 M’ misuses the word ‘used to’ and student ‘Ana Ferrer B1 M’

rephrases the sentence solving the mistake. The students ‘Marta Alo B1 M’ reflects on that mistake and agrees with herpartner. In Fig. 2 we can appreciate how the student ‘Marta Alo B1 M’ in a subsequent conversation, uses the adverb ‘usually’appropriately.

Fig. 3 shows a LRE in which the student ‘Belen B1 M’ makes a lexical mistake with the word ‘cooker’. In this case thestudent ‘Ana B B1 M’, who knows the meaning of the word her partner is trying to convey, tries to explain the meaning of theword ‘cook’. In Fig. 4, the student ‘Belen B1M’ days later in a conversation about dream jobs uses the correct form of theword.

Fig. 5 shows a LRE in which the student ‘Laura Arts B1 M’ lacks awareness of the negation in English. In this case thestudent ‘Ana Ferrer B1 M’ points out the fact that the double negation in English is not possible. The teacher supports theargument of the student, confirming that this claim is true. In Fig. 6, we can observe how the student ‘Laura Arts B1 M’ in adifferent conversation months later, uses the negation appropriately.

Fig. 1. Screenshot of a LRE elicited from the interaction in WhatsApp.

Fig. 2. Screenshot of a subsequent conversation in which the appropriate form of the word in Fig. 1 is observed.

A. Andujar / System 62 (2016) 63e7670

Fig. 3. Screenshot of a LRE elicited from the interaction in WhatsApp.

Fig. 4. Screenshot of a subsequent conversation in which the appropriate form of the word in Fig. 3 is observed.

A. Andujar / System 62 (2016) 63e76 71

Fig. 5. Screenshot of a LRE elicited from the interaction in WhatsApp.

Fig. 6. Screenshot of a subsequent conversation in which the appropriate use of negation is observed.

A. Andujar / System 62 (2016) 63e7672

A. Andujar / System 62 (2016) 63e76 73

5. Discussion

The present study was designed to determine if the programme implementation in the experimental group involvedgreater improvements in writing skills. Returning to the research questions previously posed, as shown in the group com-parison carried out in Tables 1 and 2, the ratios of lexical, grammatical and mechanical errors as well as error-free clauses perclause and error-free T-unit per T-unit indicated significant differences between the control and experimental group in termsof accuracy. Notwithstanding that MSTTR measuring syntactic complexity together with lexical diversity showed significantdifferences between the two groups, no significant differences were found in terms of syntactic complexity as the ratios ofwords per clause, words per T-unit and clauses per T-unit yielded no statistically significant difference. The control group, asshown in Table 1, experienced improvements in terms of accuracy too, which may be due to practice within the group duringthe course. Nevertheless, as it was previously mentioned, Sidak results showed significant differences in the development ofaccuracy between the two groups. These findings highlight the potential of text chat in MIM to improve second languageaccuracy, in linewith the latest studies regarding linguistic accuracy via text chat in CMC (Adams, Alwi,&Newton, 2015; Alwi,Adams, & Newton, 2012; Jepson, 2005). Figures regarding MSTTR concerning syntactic complexity together with lexicaldiversity emphasize the idea that synchronous and particularly asynchronous interaction gives learners more opportunitiesto produce syntactically complex language in linewith Sotillo's (2000) research in CMC, as students are able to reflect on theirlanguage productions. Nevertheless, the independent parameters of complexity showed no significant difference betweenthe groups studied, hence further investigation is required.

As shown in previous studies of CMC such as Bueno-Alastuey (2013), Jepson (2005) or Rouhshad, Wigglesworth, andStorch (2016), chat-based conversation is a fertile ground for LREs. In a similar vein, MIM, due to the similarities to CMC,contains many of the LREs which contribute to construct knowledge as well as to second language development. The dif-ferences in the LREs between MIM and CMC remain as an object of study as the characteristics of mobile devices offer anenvironment full of opportunities for interaction and feedback.

Regarding the theoretical implications of these findings, the combination of error perception and negotiation of meaningskills has a beneficial effect on writing development. Knowledge is built in an active manner by students, who perform a keyrole in the process paying attention not only to the feedback provided by the teacher but also to their language productions. Arich input environment is created where instead of only one source of input, all participants take part in its creation, satisfyingreal communicative needs (Rixon, 2001). Cooperative learning taking place in the activity becomes an essential source ofknowledge, and is an indispensable factor to encourage students. As Slavin (1983) claims, this kind of learning is conceived ofas a team process in which the members help each other and trust in themselves to reach a goal. Some of these goals are:developing and sharing a common project; contributing to problem-solving, through questions, reflections and solutions;encouraging participation among students.

Long's (1983) interaction theory e although it addresses communicative oral skill e is applicable in this study as thesimilarities between MIM and oral language correlate highly: spontaneity; high degree of interaction; naturalness; infor-mality. Interaction, as opposed to studies such as Levy and Kennedy (2005) - where messages were sent at set times, on setdays - takes place in a synchronous and asynchronous manner, exploiting the functionalities and advantages of MALL.

Students undergo improvements in terms of syntactic complexity together with lexical diversity and accuracy, whichcorrelate with can-do descriptors established by CEFR such as the possibility of writing about experiences, feelings, reactionsand a detailed description on a range of familiar subjects as students are able to elicit a wider range of words than in previouslevels. Regarding grammatical and mechanical skills, apart from the ones previously mentioned, we can observe severalcorrelations with other descriptors set by CEFR such as the possibility of correcting mix-ups with tenses or expressions thatlead to misunderstandings, or the ability to start again if communication breaks down. Furthermore, the possibility ofnarrating a story, writing a short or simple essay or summarising an opinion, would not be possible without such skills.

Despite the fact that the researcher in this inquiry tried to avoid all the possible limitations, some were found throughoutthe study. Participationwas a major concern and although the acceptance of the activity was remarkable with a total numberof 13,937 messages sent and a participation of 97.5% of the students, the amount of time students spent chatting through theapplicationwas not balanced as some limited themselves to the daily answers whereas others constantly followed the threadof the conversation. As such, the effects were more salient in certain students who participated to a higher degree.

6. Conclusion

Overall, the study strengthens the idea that the use of MIM contributes to the development of accuracy in second languagewriting. The significant decrease in the number of errors in the experimental group in relation to the control, whethergrammatical, lexical or mechanical, illustrates the positive effects of the treatment in terms of linguistic accuracy. Regardingcomplexity and lexical diversity, the MSTTR emphasizes greater syntactic complexity together with lexical diversity.Nevertheless, the independent parameters for syntactic complexity yielded no statistically significant difference, hence thereis no strong evidence for the MIM exchanges themselves to be significant. This may be due to its simplified register and shortlength, therefore further studies are required. In addition, the LREs taking place in the application constitute a fundamentalsource of knowledge for second language development, as students reflect and construct knowledge in an active manner,providing feedback and negotiating meaning.

A. Andujar / System 62 (2016) 63e7674

Leaving aside those questions related to learning that were previously detailed in the discussion and throughout ourresearch, the acceptance earned by the use of WhatsApp among students needs to be emphasized. According to Coll (2004),referring to didactics, the interest in studying the impact of new information and communication technologies (TIC) ineducative processes has grown progressively during the last years, in parallel with an increasing inclusion of technologies atall levels of teaching. This makes us wonder why the tremendous potential that theWhatsApp application has proved to haveas an educational tool during this research study has gone unnoticed in other studies. MALL encourages collaboration and co-construction of knowledge where learners had to find information and share it with their peers (Kukulska-Hulme & Shield,2007). In this manner, the high involvement of the students in the activity and throughout the process was remarkable, evenwhen the teacher was not participating actively. Moreover, the activity helped reduce the teacher-student distance, makingthe students feel more comfortable and increasing their participation and interest in this field of knowledge, thus facilitatinglearning. The number of studies related to MIM in education is extremely scarce and the tremendous potential of this tool hasnot been taken into consideration, thus a new field of research is open, where education and technology must progress inunison.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Acknowledgements

This research was carried out at the University of Almeria Language School, Spain, as part of a B1 English language coursein 2013e2014. I wish to thank professor Phil Benson who provided insight and expertise that greatly assisted theinvestigation.

References

Adams, R. (2003). L2 output, reformulation and noticing: Implications for IL development. Language Teaching Research, 7, 347e376.Adams, R., Alwi, N. A. N. M., & Newton, J. (2015). Task complexity effects on the complexity and accuracy of writing via text chat. Journal of Second Language

Writing, 29, 64e81.Alexander, B. (2004). Going nomadic: Mobile learning in higher education. Educause Review, 39(5), 29e35.Alwi, N. A. N. M., Adams, R., & Newton, J. (2012). Writing to learn via text chat: Task implementation and focus on form. Journal of Second Language Writing,

21(1), 23e39.American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. Retrieved from http://www.apa.org/ethics/code2002.

html.Blake, R. (2000). Computer-mediated communication: A window on L2 Spanish interlanguage. Language Learning & Technology, 4, 120e136.Bonk, C. J., & Kim, K. A. (1998). Extending sociocultural theory to adult learning. In M. C. Smith, & T. Pourchot (Eds.), Adult learning and development:

Perspectives from educational psychology (pp. 67e88). Mahwah, NJ: Erlbaum.Bueno-Alastuey, M. C. (2013). Interactional feedback in synchronous voice-based computer mediated communication: Effect of dyad. System, 41(3),

543e559.Burston, J. (2014). The reality of MALL: Still on the fringes. CALICO Journal, 31(1), 103.Campbell, D., & Stanley, J. (1963). Experimental and quasi-experimental designs for research. In N. L. Gage (Ed.), Handbook of research on teaching (pp.

171e246). Chicago, IL: Rand-McNally.Canale, M. (1983). From communicative competence to communicative language pedagogy. Language and communication, 1, 1e47.Canale, M., & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied linguistics, 1(1), 1e47.Chen, J., Belkada, S., & Okamoto, T. (2004). How a Web-based course facilitates acquisition of English for academic purposes. Language Learning & Tech-

nology, 8, 33e49.Cobcroft, R. S., Towers, S. J., Smith, J. E., & Bruns, A. (2006). Mobile learning in review: Opportunities and challenges for learners, teachers and institutions. In

Proceedings at online learning and teaching (OLT) conference. Brisbane, Australia. Retrieved from http://eprints.qut.edu.au/5399/1/5399.pdf.Cohen, A. D. (1983). Reformulating second-language compositions: A potential source of input for the learner. Retrieved from ERIC database. (ED228866).Coll, C. (2004). Psicología de la educaci�on y pr�acticas educativas mediadas por las tecnologías de la informaci�on y la comunicaci�on. [Psychology of education

and educational practices mediated by information technology and communication]. Sin�ectica, 25, 1e24.DeKeyser, R. M., & Larson-Hall, J. (2005). What does the critical period really mean? In J. F. Kroll, & A. M. B. de Groot (Eds.), Handbook of bilingualism:

Psycholinguistic approaches (pp. 88e108). Oxford, UK: Oxford University Press.Desai, C., & Graves, S. (2006). Instruction via instant messaging reference. The Electronic Library, 24, 174e189.Diaz, J. (2014). El WhatsApp como herramienta de intervenci�on did�actica para fomentar el aprendizaje cooperativo [WhatsApp as a tool of educational inter-

vention to foster cooperative learning]. Unpublished Paper presented at X Jornadas De Material Did�actico y Experiencias Innovadoras En Educaci�onSuperior, Buenos Aires, Argentina. Retrieved from http://eprints.rclis.org/23597/.

Dobao, A. F. (2012). Collaborative writing tasks in the L2 classroom: Comparing group, pair, and individual work. Journal of Second Language Writing, 21,40e58.

Doughty, C. J., & Long, M. H. (2003). Optimal psycholinguistic environments for distance foreign language learning. Language Learning & Technology, 7,50e80.

Dourando, D., Parker, M. D., & De la Harpe, R. (2007). In An investigation into the usage of mobile instant messaging in tertiary education. Proceedings of AnnualConference of World Wide Web Application. Johannesburg, South Africa. Retrieved from http://marlonparker.co.za/dourandoparker.doc.

Farmer, R. (2003). Instant messaging: Collaborative tool or educator's nightmare!. Retrieved from https://www.student.cs.uwaterloo.ca/~cs492/06public_html/papers/im.pdf.

Gass, S. (2003). Input and interaction. In C. Doughty, & M. Long (Eds.), The handbook of second language acquisition (pp. 224e255). Oxford: Blackwell.Gülbahar, Y., Jacobs, C., & K€onig, A. (2015). Mobile learning in higher education. International handbook of e-learning volume 2: Implementation and case studies

(pp. 33e42). New York, NY: Routledge.Hanaoka, O., & Izumi, S. (2012). Noticing and uptake: Addressing pre-articulated covert problems in L2 writing. Journal of Second Language Writing, 21,

332e347.Hymes, D. H. (1972). On communicative competence. In J. B. Pride, & J. Holmes (Eds.), Sociolinguistics. Harmondsworth, England: Penguin Books.

A. Andujar / System 62 (2016) 63e76 75

Izumi, S. (2002). Output, input enhancement, and the noticing hypothesis. Studies in Second Language Acquisition, 24, 541e577.Izumi, S., & Bigelow, M. (2000). Does output promote noticing and second language acquisition? TESOL Quarterly, 34, 239e278.Izumi, S., Bigelow, M., Fujiwara, M., & Fearnow, S. (1999). Testing the output hypothesis. Studies in Second Language Acquisition, 21, 421e452.Jepson, K. (2005). Conversation and negotiated interaction in text and voice chat rooms. Language Learning & Technology, 9, 79e98.Koole, M. (2009). A model for framing mobile learning. In M. Ally (Ed.), Mobile learning: Transforming the delivery of education and training (pp. 25e47).

Edmonton, Canada: AU Press.Kukulska-Hulme, A., & Shield. (2007). An Overview of Mobile Assisted Language Learning: Can mobile devices support collaborative practice in speaking

and listening. In EuroCALL, 2007.Lantolf, J., & Thorne, S. (2006). Sociocultural theory and the genesis of second language development. Oxford, UK: Oxford University Press.Lapkin, S., & Swain, M. (2004). What underlies immersion students' production: The case of avoir besoin de. Foreign Language Annals, 37, 349e355.Lapkin, S., Swain, M., & Smith, M. (2002). Reformulation and the learning of French pronominal verbs in a Canadian French immersion context. The Modern

Language Journal, 86, 485e507.Levy, M., & Kennedy, C. (2005). Learning Italian via mobile SMS. In A. Kukulska-Hulme, & J. Traxler (Eds.), Mobile learning: A handbook for educators and

trainers (pp. 76e83). London, UK: Taylor and Francis.Long, M. H. (1983). Classroom-oriented research on second language acquisition. Rowley, MA: Newbury House.Mayo, E. (1933). The human problems of an industrial civilization. New York: Macmillan.Morat�o Pay�a, A. (2014). El WhatsApp como complemento del aprendizaje en la clase E/LE [WhatsApp as a learning complement in the classroom of E/LE].

Revista Foro De Profesores De E/LE, 10(1), 165e173.Muniandy, A. V. A. (2002). Electronic-discourse (E-discourse): Spoken, written or a new hybrid? Prospect: An Australian Journal of TESOL, 17, 45e68.Negretti, R. (1999). Web-based activities and SLA: A conversation analysis research approach. Language Learning & Technology, 3, 75e87.Nunan, D. (1992). Research methods in language learning. Cambridge: Cambridge University Press.Ohta, A. S. (2000). Rethinking interaction in SLA: Developmentally appropriate assistance in the zone of proximal development and the acquisition L2

grammar. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 51e78). Oxford: Oxford University Press.Ohta, A. S. (2001). Second language acquisition processes in the classroom: Learning Japanese. Mahwah, NJ: Lawrence Erlbaum.Ortega, L. (1997). Processes and outcomes in networked classroom interaction: Defining the research agenda for L2 computer-assisted classroom discussion.

Language Learning & Technology, 1, 82e93.Oxford, R. (2001). Integrated skills in the ESL/EFL classroom. Washington DC: Clearinghouse on Languages and Linguistics.Padron, C. J. (2013). Teaching strategies based on WhatsApp instant messaging application only for phones (mobile learning) and its use to promote

collaborative learning. Revista De Tecnología De Informaci�on y Comunicaci�on En Educaci�on, 7(2), 123e134.Payne, J., & Whitney, P. J. (2002). Developing L2 oral proficiency through synchronous CMC: Output, working memory, and interlanguage development.

CALICO Journal, 20, 7e32.Pellettieri, J. (1996). Network-based computer interaction and the negotiation of meaning in the virtual foreign language classroom. Unpublished manuscript.

University of California at Davis.Pellettieri, J. (2000). Negotiation in cyberspace: The role of chatting in the development of grammatical competence. In M. Warschauer, & R. Kerns (Eds.),

Network-based language teaching: Concepts and practice (pp. 59e86). Cambridge: Cambridge University Press.Pica, T. (1994). Research on negotiation: What does it reveal about second-language learning conditions, processes, and outcomes? Language Learning, 44,

493e527.Rambe, P., & Bere, A. (2013). Using mobile instant messaging to leverage learner participation and transform pedagogy at a South African University of

Technology. British Journal of Educational Technology, 44, 544e549.Rixon, S. (2001). Optimum age or optimum conditions? Issues related to the teaching of languages to primary age children. Retrieved from http://www.

britishcouncil.org/english/eyl/article01.htmhttp://www.britishcouncil.org/english/eyl/article01.htm.Rouhshad, A., Wigglesworth, G., & Storch, N. (2016). The nature of negotiations in face-to-face versus computer-mediated communication in pair in-

teractions. Language Teaching Research, 20(4), 514e534.Schmidt, R. (1995). Consciousness and foreign language learning: A tutorial on the role of attention and awareness in learning. Honolulu, HI: Second Language

Teaching and Curriculum Center, University of Hawaii.Shudong, W., & Higgins, M. (2006). In Limitations of mobile phone learning. IEEE international workshop on wireless and mobile technologies in education

(WMTE), Tokushima, Japan, 2005. Retrieved from http://ieeexplore.ieee.org/xpls/icp.jsp?arnumber¼1579260.Skehan, P. (2009). Modelling second language performance: integrating complexity, accuracy, fluency, and lexis. Applied linguistics, 30(4), 510e532.Slavin, R. E. (1983). When does cooperative learning increase student achievement? Psychological Bulletin, 94, 429e445.Smith, B. (2003). Computer-mediated negotiated interaction: An expanded model. The Modern Language Journal, 87, 38e54.Smith, B. (2004). Computer-mediated negotiated interaction and lexical acquisition. Studies in Second Language Acquisition, 26, 365e398.Sotillo, S. (2000). Discourse functions and syntactic complexity in synchronous and asynchronous communication. Language Learning & Technology, 4,

82e119.St. John, E., & Cash, D. (1995). Language learning via e-mail: Demonstrable success with German. In M. Warschauer (Ed.), Virtual connections: Online activities

and projects for networking language learners (pp. 191e197). Honolulu, HI: University of Hawaii.Storch, N. (2002). Patterns of interaction in ESL pair work. Language Learning, 52, 119e158.Storch, N., & Wigglesworth, G. (2007). Writing tasks: Comparing individual and collaborative writing. In M. P. García Mayo (Ed.), Investigating tasks in formal

language learning (pp. 157e177). London: Multilingual Matters.Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. Gass, & C. G.

Madden (Eds.), Input and second language acquisition (pp. 235e253). Rowley, MA: Newbury House.Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty, & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp.

64e81). Cambridge: Cambridge University Press.Swain, M. (2006). Languaging, agency and collaboration in advanced second language proficiency. In H. Byrnes (Ed.), Advanced language learning: The

contribution of Halliday and Vygotsky (pp. 95e108). New York, NY: Continuum.Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied Linguistics,

16, 371e391.Swain, M., & Lapkin, S. (1998). Interaction and second language learning: Two adolescent French immersion students working together. The Modern

Language Journal, 82, 320e337.Swain, M., & Watanabe, Y. (2012). Languaging: Collaborative dialogue as a source of second language learning. In C. Chapelle (Ed.), The encyclopedia of

applied linguistics. Oxford: Wiley-Blackwell.Thornbury, S. (1997). Reformulation and reconstruction: Tasks that promote ‘noticing’. ELT Journal, 51, 326e335.Thorndike, E. L. (1911). Animal intelligence: Experimental studies. New York, NY: Macmillan.Trifonova, A. (2003). Mobile learning-review of the literature (Technical report #DIT-03-009). Italy: Department of Information and Communication Tech-

nology, University of Trento.V�azquez-Cano, E., Mengual-Andr�es, S., & Roig-Vila, R. (2015). An�alisis lexicom�etrico de la especificidad de la escritura digital del adolescente en Whatsapp.

[Lexicometric analysis of the specificity of teenagers' digital writing in whatsapp]. Revista de Lingüística Te�orica y Aplicada, 53(1), 83e105.Vygotsky, L. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.Warner, C. N. (2004). It's just a game right? Types of play in foreign language CMC. Language Learning & Technology, 8, 69e87.Warschauer, M. (1996). Comparing face-to-face and electronic discussion in the second language classroom. CALICO Journal, 13, 7e26.

A. Andujar / System 62 (2016) 63e7676

Werry, C. C. (1996). Linguistic and interactional features of internet relay chat. In S. Herring (Ed.), Computer-mediated communication: Linguistic, social, andcross-cultural perspectives (pp. 47e63). Amsterdam, The Netherlands: J. Benjamins.

Wigglesworth, G., & Storch, N. (2009). Pair versus individual writing: Effects on fluency, complexity and accuracy. Language Testing, 26, 445e466.Yang, L., & Zhang, L. (2010). Exploring the role of reformulations and a model text in EFL students' writing performance. Language Teaching Research, 14,

464e484.