The Role Of Interactional Modification In Efl

The interaction hypothesis claims that interactional modification promotes language development through increasing comprehension. Owing to the fact that online interactive exchange offers learners many opportunities to use the target language to negotiate both meaning and form in a social context , this study is an attempt to compare the types and frequency of Interactional Modifications as employed by two groups of intermediate EFL students-one interacting via online communication and the other through regular oral class interaction– to explore the potentials of text-based online chat in facilitating the development of question forms in English as a foreign language. To do so, while a total of eighteen paired intermediate EFL students used a synchronous chat program (Yahoo Messenger) over a seven week period to complete a series of communicative tasks outside their classroom time, the members of the comparison group did the same tasks using plain classroom oral interaction. The results show that Iranian EFL learners in computer mediated communication (CMC) environment had over five times as many turns, and used a much greater variety of interactional modifications in comparison to the group in class environment. Data from this study also suggest that interactional modifications during online negotiation facilitate the development of question forms in English as a foreign language. Given that the current study only focused on online interaction among non- native speakers (NNSs), future studies on online negotiation including both NNSs and native speakers (NSs) are still needed.

Key terms: Computer-Mediated Communication, Interactional modifications, question forms development

Introduction

Interaction is the key to second language learning. However one of the greatest challenges faced by the foreign language (FL) education is how to construct an interactive learning environment outside the classroom in which learners can exchange information and communicate ideas in the target language (TL). This challenge as Campbell (2004) truly mentioned is due to the fact that unlike ESL learners who communicate in TL outside of the classroom, the EFL learners “re-enter a world” talking their mother tongue as soon as they leave the classroom and consequently they don’t have any opportunities to interact in TL and are left with “little opportunity to use what they’ve learned” in the classroom.

Many a researcher in the area of Computer Assisted Language Learning (CALL) has suggested that text-based Synchronous Computer Mediated Communication (SCMC) is capable of providing an ideal learning environment which may be beneficial to second/foreign language learning. (Beauvois, 1992; Pellettieri, 1999; Chapelle, 2001). Pellettieri (2000) declared a logical relationship and argued that

because oral interaction is considered by many to be important for second language development, and because Synchronous [CMC]. . . bears a striking resemblance to oral interaction, it seems logical to assume that language practice through [CMC] will reap some of the same benefits for second language development as practice through oral interaction.(p. )

While many studies investigated the role of oral negotiated interaction in second/foreign language development (Gass, 1997; Long, 1985; Mackay, 1995; Mackey & Philp, 1998), the role of online negotiated interaction in facilitating language development has not yet been explored completely. Specifically, as far as the efficacy of SCMC medium in grammatical development is concerned, the previous findings are unsatisfactory and controversial. Moreover little research has yet been conducted into the effect of CMC in the EFL context. This study therefore, explores the role of interactional modifications during text-based SCMC in order to better understand the potential of text-based online chatting to facilitate the development of question formation in English as a foreign language.

Literature review

Interactional modifications in CMC environment

In spite of the fact that a few studies has addressed the issue of interactional modifications in CMC environment (Lee, 2001, 2002; Kotter, 2003; Jepson 2005; Isharyanti, 2008), none of these studies have been conducted in naturalistic setting (e.g. outside of the classroom or language laboratory), thus ignoring the real potential of CMC technology which is beyond any temporal and spatial constraints. Although focusing on different categorization of interactional modifications, all of these studies have shown unanimously that CMC medium is capable of providing a conductive learning environment in which interactional modifications might be generated.

Lee (2001, 2002) conducted two similar studies in which she explored the types of interactional modifications employed by Spanish Learners at intermediate level of proficiency. Considering both studies, a total of ten categories of interactional modifications were identified including Comprehension checks; Clarification checks; Confirmation checks; Use of English, Word invention; Request (for help); Use of approximation; Self corrections; Topic shift ,and Use of keyboard symbols as discourse makers. The result showed that request (for help) happened the most frequently followed by clarification checks self- correction and comprehension check (Lee, 2001, p. 238; Lee, 2002, p.280). Through a comparison of the result of her study with the literature on face-to-face communication Lee (2002) argued that the learners negotiated with each other “using a variety of modification devices similar to face-to face communication” (p. 280). Being one of the first studies in this area, the studies conducted by Lee were subject to a number of limitations. For example the definitions used in coding categories are somewhat problematic (Thomas & Reinders, 2010) since as Kotter (2003) stated “there is substantial overlap between the definitions of clarification checks and requests” (p. 157).

Providing a “more coherent classificatory system”, the study by Kotter (2003) focused on eight types of interactional modifications including: Confirmation checks; Clarification request; Comprehension checks, Repetitions; Recasts, Overt indications of understanding; Overt indications of agreement; and Overt indications of non-agreement (p. 157). However, in contrast to Lee’s (2001, 2002) finding, Kotter’s (2003) study revealed a “marked difference between conversational repair in spoken interactions and in the MOO-based exchange” (p. 145). He discussed that these differences may due to a number of “medium-specific” factors (p.163).

Jepson (2005) explored the types of repair moves used by NNSs in synchronous text-based chat in comparison to voice-based chat during 10, 5-minute sessions. (5 text-based chat sessions and 5 voice-based chat sessions). Although a number of new interactional modifications ( e.g. Self repetition/Paraphrase, Explicit Correction and Question)were identified, this study is also subject to a number of limitations. There was no information about the proficiency level of participants as this issue may affect the interaction and the interactional modifications generated during the process of negotiation. Furthermore, the duration of the sessions was limited (5 minutes). These two issues may question the external validity of the findings.

In a more recent study, Isharyanti (in Marriott & Torres, 2008) examined the types and frequency of interactional modifications employed by NNSs during synchronous text- based chat. Focusing on a more varied number of interactional modifications, Isharyanti’s study confirmed the potential of CMC environment in generating interactional modifications. The data showed that the participants engaged in online negotiation process and in order to understand and to be understood, they used a variety of interactional modifications among which “confirmation check” (24%) was the most frequent one followed by “overt indication of agreement” (21%) and “clarification request” (21%). Nevertheless, to examine the potential of CMC technology as a medium for negotiated interaction, the presence of a control a group (a typical class environment in which learners communicate face-to-face) would be helpful.

L2 grammatical development and Text-based computer mediated communication

The role of negotiated interaction in L2 development continues to be of great interest to researchers (see Pica, 1987; Gass and Varonis, 1989; Mackey, 1995). However, much of the research on the effect of negotiated interaction and interactional modifications used during negotiation process is based on data from face-to-face interaction in ESL context. Moreover, the few studies that have been conducted in CMC environment are unsatisfactory and controversial. There are three stand…………………….blue print

Sotillo (2000) compared syntactic complexity of 25 learners’ output produced in synchronous versus asynchronous computer mediated communication (SCMC vs. ACMC). She claimed that because of the fast nature of interaction in SCMC, the participants did not pay any attention to form (accuracy) while negotiating meaning (97). In the meantime, due to the delayed nature of ACMC, this medium was capable of generating more syntactically complex structures. Nevertheless, as Fiori (2005) stated “while Sotillo reported that grammatical accuracy may suffer in the SCMC environment, her data revealed that the synchronous group’s interactions exhibited fewer errors than the asynchronous group’s utterances” (p. 569). Further investigation, therefore is needed to better understand the potential of SCMC environment in fostering the grammatical development.

In the same vein of research, Lee’s studies (2001, 2002) admitted that although the participants did engage in negotiation of meaning, negotiation of form rarely happened among them. She argued that due to the rapidity of the interaction occurred in SCMC, participants produced brief utterances using simple sentence structures and ignored linguistic errors (Lee, 2001, p. 239). Thus she concluded that interaction mediated via SCMC technology fostered fluency rather than accuracy. Emphasizing on further studies on the effectiveness of CMC medium for the development of learners’ interlanguage, Lee (2001) suggested that “students need to be advised of the need to write correctly to maintain a balance between function, content, and accuracy” (p. 242).

Blake (2000) examined the online discourse of 50 Spanish learners at intermediate level of proficiency in term of task type. By considering the potential of CMC environment for eliciting negotiation of meaning, Blake (2000) concluded that “carefully crafted tasks stimulate L2 learners to negotiate meaning which. . . . appear to constitute ideal conditions for SLA, with the CMC medium being no exception” (p. 133). However, his stand on the effectiveness of CMC medium for grammatical development is uncertain. According to Blake (2000) “the predominance of incidental lexical negotiations, in contrast to the paucity of syntactic negotiations, leaves unanswered or unsatisfactorily addressed the issue of grammatical development” (p. 120). Therefore, further research is to be conducted in order to declare this issue with certainty.

On the contrary, Pellettieri (2000) claimed that bearing a “striking resemblance” to oral interaction, “network based communication (NBC)” is capable of facilitating grammatical development. She examined online discourse produced by 20 NNSs during 5 30-minute sessions. As far as the potential of CMC medium in stimulating negotiation of meaning is concerned, the results of her study echoed the findings reported in literature. She observed that, engaging in online negotiation process is as facilitative as it is typical to oral interaction. She also found that the learners seek mutual understanding and try to “convey their meaning” by using some communicative strategies. Adopting an interactionist perspective, she argued that

Because through negotiation, interlocutors can zero in on the exact source of communicative problem they are trying to resolve, and because often at the root of the problem is some aspect of the L2 form, be it lexical, syntactic, or semantic, L2 learners are even more likely to notice the problem and attend to these very aspects of form in their output while negotiating meaning. (p.61)

As Pellettieri reasoned, engaging in the process of negotiation of meaning, the interlocutors’ attention will be drawn on linguistic form and consequently they are able to notice the gap between their interlanguage and the target form. This process “focus on form” has been claimed necessary for grammatical development (Gass and Varonis, 1994; Schmidt, 1990; Spada & Lightbown, 1993 as cited in Pellettieri, 2000).

Finally Salaberry (2000) compared the efficacy of two environments: CMC environment versus face-to-face environment in fostering “L2 morphosyntactic development”. He analyzed the discourse produced by four NNSs qualitatively. The findings revealed that “with respect to the use of past-tense verbal morphology across tasks, some initial changes in the development of morphological endings were more evident in the CMC session than in oral session” (p. 17). According to Salaberry (2000), the outperforming of CMC environment over face-to-face environment in facilitating morphological development may due to medium-specific characteristic that is CMC provides a learning environment in which the learners’ attention will be drawn on both form and function (p. 19).

Differences in the mixed findings regarding the effect of SCMC medium on the grammatical development may be due to the difference in target items used to measure development. In Sotillo’s (2000) study, the target feature was “Syntactic complexity” which was defined by Sotillo’s (2000) as “the ability to produce writing that uses subordination and embedded subordinate clauses” (p. 99). Lee’s (2001, 2002) measure of development was sentence structure (simple sentences versus complex ones) ,while Salaberry (2000) used “past tense verbal endings” as the target grammatical feature. However, as far as the researchers are aware, there is not a single study to address the effect of CMC environment on the development of grammatical development with regard question forms in English as a foreign language. Therefore, in the light of ongoing discussion as to the role of CMC in foreign language development, the present study was designed to answer the following research questions:

Do Iranian EFL learners engage in meaning negotiation process through SCMC?

What types and frequency of interactional modifications do Iranian learners employ in CMC environment versus of those in class environment?

Is there a the relationship between group membership (Experimental group: CMC environment Vs. Control group: Class environment) and the production and development of question forms in English as a foreign language?

Method

Following the mixed methods approach, this study adopted both a qualitative perspective– to provide an in-depth explanation of the types of interactional modifications– and a quantitative view point via a pretest-posttest, delayed posttest design — to explore the existing relationships.

Participants

A total of 36 EFL learners from a Language school in Rasht, Iran participated in the present study. Their Participation in the study was voluntary and involved a commitment of 50 to 60 minutes performing some communicative tasks with their assigned partners for one session per week for 7 weeks outside of the classroom. This study was held during study periods at the language school. All participants were native speakers of Persian who were receiving two 90-minute classes of English instruction weekly. There were 30 female and 6 male students. The age of participants ranged from 14 to 32 years, with an average of 17. To meet local ethics requirements, the participants’ and their parents consent to contribute to this project was secured through two forms– a Contract Form with an explanatory statement according to which the researchers guaranteed that participants would receive some rewards (some English books on DVD) if they accomplished the expected tasks and Parental Consent Form, in Persian, given to all participants and their parents to be signed.

Instrument

Yahoo! Messenger Chat Software

The software used in this study was the Yahoo! Messenger, a free program available for public use. It allows for real time, synchronous Computer Mediated Communication in Internet chat rooms. Using the “text mode”, the participants could record all of the written transactions entered in a chat window, which provided the researchers with an instantaneous transcript of all user exchanges.

Communication Tasks

The tasks used in this study were selected and developed to (a) provide opportunities for the interactional modifications to take place and (b) provide context for the targeted structures to occur. Following Pica, Kanagy & Falodun’s typology (1993), information gap tasks were used in this study in order to make the participants exchange information and endeavor to gain a single outcome. The selection of the communicative tasks –utilized for both treatment and tests-was motivated by previous studies such as Pellettieri, (1999), Blake, (2000) and Cheon (2003).

Each task was photocopied and distributed to every participant. Table 1. presents a detailed description of each task.

Table 1. Task materials used for test and treatment

Task

Task instruction given to participants

Pretest

Missing information

Work with your partner. Ask and answer questions to find the missing information.

Treatment

Complete the drawing

Below is a drawing of Richard’s room. He hasn’t had time to put all his things where he wants them. Your partner has a complete drawing of his room. Ask him/her questions where to put all the things.

Complete the drawing

Below is a drawing of a kitchen. Your partner has the same drawing but with a number of objects. (E.g. glasses, pots, etc.). Ask him/her questions where to put all the things.

Spot the differences

You both have two similar photographs but taken at a slightly different time. Work with your partner to find as many differences between the two photos as you can.

Spot the differences

Work with your partner. You both have a drawing of a busy yard where you can see people doing different things. Your drawings are NOT the same. There are 9 differences. Ask and answer questions to find the differences.

Posttest

Missing information

Work with your partner. Ask and answer questions to find the missing information.

Delayed Posttest

Missing information

Work with your partner. Ask and answer questions to find the missing information.

Procedures

First, the researchers explained the project to the students and expressed the hope that the students would choose to participate. All participation was voluntary, however to encourage the students to take part in the study, the instructor guaranteed that participants would receive a DVD-pack including some English books as a reward. To meet local ethics requirements, through a Contract Form with an explanatory statement and a Parental Consent Form in Persian, the students were asked if they would like to participate in this research project. Afterwards, Participants in experimental group (n=18) were divided by self-selecting into nine pairs and were asked to chat online to do a series of communicative tasks for one session per week for 7 weeks outside of the classroom. The students completed 7 tasks in total over the duration of the study. There were no time limits imposed on tasks. And the students were told that they should send copies of their chat transcripts to the researchers’ e-mail address for further data analysis. The participants in control group (n=18), solved the same tasks face-to-face in the classroom within the limited time imposed by …

Data collection and analysis

The data were collected from students’ weekly online chatting together with the scripts from face-to face conversation. In order to investigate whether Iranian learners engage in meaning negotiation through SCMC technology, the data collected from the written discourse were analyzed based on the typical scheme established by Varonis and Gass (1985). In accordance with Blake’s (2000) study, the number of total turns and negotiations was calculated. Accordingly, negotiation routines were identified by means of their four components: trigger, indicator, response and reaction. The example below drawn from this study illustrates the model during lexical negotiation.

Student A: It’s on the drawer, near the edge of it. [Trigger]

Student B: what do you mean by edge? [Indicator]

Student A: side or lip. [Response]

Student B: ok. Thanks. [Reaction]

A ratio of negotiated turns to total turns was calculated in order to make the data comparable with previous studies. Additionally, to identify the type and frequency of interactional modifications, the data were coded based on the thirteen different interactional modifications using the criteria provided in Table 2. Using a selection of interactional modifications from a number of studies on internet chatting (Lee, 2001, 2002; Kotter, 2003; Jepson, 2005), the categorization of interactional modifications in Isharyanti, (n.d.) in Marriott & Torres (2008) motivated the categorization system in this study. There were a number of modifications. Three items namely “Explicit correction”, “use of Persian”, and “question” were added to the aforementioned category. In order to establish inter-coder agreement, a trained independent coder recoded a randomly-selected 25% of the data. A Pearson product-moment correlation revealed an inter-rater correlation of .85 (P<0.001), suggesting that the data were coded with strong consistency.

Table 2. Categories, definition and example of interactional modifications

Code

Type of IM

Definition

Example

1

CC

Confirmation check

A speaker’s attempt to confirm that he has understood an utterance via the (partial) paraphrase (as opposed to repetition, see below) of this turn, which can simply be answered with Yes or No.

Did you mean?

2

CR

Clarification request

An explicit demand for an elaboration or a reformulation of an idea, which “requires a rerun of the troublesome utterance” in question.

What do you mean by X?

3

C

Comprehension check

A speaker’s attempt to prompt another speaker to acknowledge that he has understood a particular utterance.

Do you Understand?

4

SR/P

Self Repetition

The repetition, in isolation, of part of or an entire erroneous or otherwise problematic utterance.

Where is the lamp?(*2)

5

SC

Self-correction

To correct errors made on lexical items or grammatical structure.

This has been bee, I mean been.

6

R

Recast

A form-focused partner-related target-like reformulation of all or part of an incorrect utterance.

A: I live Iran.

B: Do you really live in Iran?

7

EC

Explicit correction

You should say X.

8

InU

Overt indication of understanding

An overt indication that a speaker has understood a particular message.

Ok, I got it thanks.

9

InA

Over indication of agreement

An overt indication that a speaker agrees with what his partner said.

Yes, I agree, you’re right.

10

InNA

Overt indication of non-agreement

An overt indication that a speaker does not agree with what his partner said.

No, I think choice B is better for him.

11

P

Use of Persian

To use Persian to substitute words or ideas in English.

“second tabagheh”

12

KS

Use of keyboard symbols as discourse

markers

To signal for uncertainty or to confirm an idea or agreement.

☺

13

Q

Question

Interlocutor asks a question in order to prompt the speaker to make a question

Can you try that again?

Note: Adopted from the categorization of Interactional Modifications described in Isharyanti (in Marriott & Torres, 2008). There were a number of modifications to this classification as mentioned before.

To answer the third research question-to examine whether there is a relationship between group membership (CMC environment Vs. Classroom environment) and development in question formation–all the question forms produced by the participants during online negotiation and face-to-face interactions were collected for detailed study. Question development in this study is understood from two perspectives: (1) accuracy perspective and

(2) developmental stage perspective.

In analyzing accuracy, because participants took considerably greater time in online chat, and thus their language output in these two modes could have been different, the researchers converted the number of well-formed questions into standardized scores by computing the ratio of each participant’s number of well-formed questions to the total number of questions he or she produced. These accuracy-based scores were examined for further investigation.

In addition to accuracy, the analysis concentrated on developmental stages. All the questions, produced by the participants, were coded in terms of the developmental stage it represented based on categorization of the “developmental stages” established by Pienemann et al. (1987), shown in Table 3. Accordingly, each student’s questions were studied individually and assigned to the appropriate stage category in each task. Stage assignment was based on the highest stage from which a participant produced two linguistically unique questions. Question development was operationalized as a stage increase on either posttest or delayed posttest. It was then possible to examine whether the learners had improved over time.

Table 3. Examples of Question Forms and Developmental Stages

Stage

Description of stage

Examples

2

SVO?

Canonical word order with question intonation.

It’s a monster?

Your car is black?

You have a cat?

I draw a house here?

3

Fronting: Wh/Do/Q-word

Direct questions with main verbs and some form of fronting.

Where the cats are?

What the cat doing in your picture?

Do you have an animal?

Does in this picture there is a cat?

4

Pseudo Inversion: Y/N, Cop.

In Y/N questions an auxiliary or modal is in sentence initial position.

In Wh-questions the copula and the subject change positions.

(Y/N) Have you got a dog?

(Y/N) Have you drawn the cat?

(Cop) Where is the cat in your picture?

5

Do/Aux 2nd

Q-word->Aux/modal ->subj (main verb, etc.)

Auxiliary verbs and modals are placed in second position to Wh-Q’s (& Q-words) and before subject

(Applies only in main clauses/direct Q’s).

Why (Q) have (Aux) you (sub) left home?

What do you have?

Where does your cat sit?

What have you got in your picture?

6

Cancel Inv, Neg Q, Tag Qu

Cancel Inv: Wh-Q inversions are not present in relative clauses.

Neg Q: A negated form of Do/Aux is placed before the subject.

Tag Q: An Aux verb and pronoun are attached to end of main clause.

Can Inv) Can you see what the time is?

(Can Inv) Can you tell me where the cat is?

(Neg Q) Doesn’t your cat look black?

(Neg Q) Haven’t you seen a dog?

(Tag Q) It’s on the wall, isn’t it?

Note. This table is based on Pienemann and Johnston (1987) and Pienemann, Johnston, and Brindley (1988) as cited in Mackey and Philp (1998).

Results and discussion

The first research question addresses the engagement of Iranian learners in meaning negotiation through SCMC. As the participants took considerably greater time in online chat, and thus their language output in these two environments was expected to be different, it was necessary to provide an overview of the number of total turns for each environment. Table 4 shows the raw number of turns, negotiations, and relative number of negotiations to total turns for both CMC and Class environments. As shown in Table 4, online environment produced about six times more turns than the class environment (4445 vs. 778). Similarly CMC environment generated more negotiations than the class environment. The relative amount of negotiation in CMC environment (2.15 %) reveals that the learners engaged in negotiated interaction in CMC environment in comparison to class environment (.38 %), although negotiations encompassed a small part of total turns in both environment. This result echoed the findings in Blake’s study (2000) that is “the total number of negotiations comprises only a small fraction of the overall conversational turns, ranging from .3% to 3.8 %” (p. 127). Interestingly, the participants in CMC environment engaged more in negotiation process than class environment. This may due to the absence of paralinguistic and nonverbal information together with sufficient practice time which provide the participants with a learning environment to negotiate meaning.

Table 4. Total number of turns and negotiated turns in CMC environment and class environment

Negotiations

Total turns

Negotiations/ Total turns

CMC environment

96

4445

2.15 %

Class environment

3

778

.38 %

The second research question aimed to determine the type and frequency of interactional modifications in CMC environment versus Class environment. With regard to the efficacy of CMC technology in generating interactional modifications, the result of this study supported the findings in literature and admitted the potential of CMC medium in providing a conductive learning environment for Interactional modifications to occur. Synchronous online interaction did provide NNSs many opportunities to negotiate meaning using a variety of interactional modifications. According to Figure1 which presents the percentage of each IMs in both CMC and class environment, the participants engaged in online negotiation employed a much greater variety of interactional modifications in comparison with the face-to-face interaction. In CMC environment 12 types of interactional modifications occurred among which clarification request (25%), Confirmation check (15%), overt indication of understanding (14%), Use of keyboard symbols (12%) and Self Repetition/paraphrase (11%) were the most used interactional modifications for negotiation. However just 3 types of interactional modifications occurred in class environment. Clarification request (78%) occurred most frequently, followed by Confirmation check and Self repetition/Paraphrase (11%).In fact, this difference may, as Salaberry (2000) suggested, may be due to the “inherent characteristics of the discourse of text-based CMC (e.g. written mode of communication, absence of paralinguistic and nonverbal information)” (p. 9) which puts more pressure on people to “find the right words” than does engagement in face-to-face discourse. Unlike spoken discourse, where pitch, smiles, laughter and other cues are often employed sub-consciously, people engaged in written CMC must put all their ideas and actions into words if they want to share them with their partners. Moreover, the sufficient time during online communication made it possible for the participants to produce more interactional modifications.

The following unedited example shows how the participants during online negotiation process used clarification request and comprehension check in order to resolve the communicative problem.

Student A: where is the quilt?

Student B: what do you mean by quilt? [Clarification request]

Student B: I don’t understand.

Student A: it is the bed covering with soft material in it

Student A: got it? [Comprehension check]

Student B: yes. It is under the rug

The above example not only shows that participants employed a variety of interactional modifications to understand and to be understood in the negotiation process but also reveals that they worked collaboratively and help each other in order to solve the communicative task. The data from this study admitted that CMC provides a “collective scaffolding learning environment” in which the students can communicate ideas and exchange information through the social interaction (Warschauer, 1997 as cited in Lee, 2002).

Figure 1. The percentage of each IM in both environments

The third research question addresses the relationship between group membership (Experimental group: CMC environment Vs. Control group: Class environment) and the production and development of question forms in English as a foreign language. Accordingly, as the target feature, a total of 1667 questions (952 questions in CMC environment and 715 questions in class environment) were found. From the accuracy point of view, a one-way repeated measures ANOVA was conducted to compare scores on the accuracy(the ratio of each participant’s number of well-formed questions to the total number of questions he or she produced) in coping with statistics test at Time 1(pretest), Time 2(posttest) and Time 3(delayed posttest). The means and standard deviation are presented in Tables 5 & 6. There was a significant effect for time in CMC environment. [Wilk’s Lambda=. 67, F (2, 16) =3.86, p<.05, multivariate partial eta squared=.33].

Table 5. Mean number of well-formed questions in CMC environment

Time period

N

Mean

Standard deviation

Time 1 (pretest)

18

76.42

14.79

Time 2 (posttest)

18

81.49

14.47

Time 3 (delayed posttest)

18

89.12

13.89

Table6. Mean number of well-formed questions in Class environment

Time period

N

Mean

Standard deviation

Time 1 (pretest)

18

81.60

12.89

Time 2 (posttest)

18

85.90

17.21

Time 3 (delayed posttest)

18

89.96

12.69

A similar result was obtained for class environment. [Wilk’s Lambda=. 67, F (2, 16) =3.92, p<.05, multivariate partial eta squared=.33]. The results from the repeated measures in both environments revealed a significant difference between the three sets of accuracy-based scores in pretest, posttest and delayed posttest. Explanation is needed how similar they are based on the statistics and figure below Arguably this comparison confirmed Pellettieri’s (2000) study that similar to oral interaction, text-based online chat promoted the noticing of problematic linguistic structures and thus was beneficial to the development of grammatical competence. The findings in this study also supported Blake’s (2000) study that “carefully crafted tasks stimulate L2 learners to negotiate meaning which. . . . appear to constitute ideal conditions for SLA, with the CMC medium being no exception” (p. 133). However, in contrast to findings reported by Sotillo (2000) and Lee (2001, 2002) — that quick interactions in CMC environment encourage fluency rather than accuracy-this study revealed that online negotiated interaction (under special conditions) may lead to accuracy as well. This may due to the fact that the participants in this study as suggested by Lee (2001) herself, were advised to reexamine and revise their chat transcript and were encouraged to write correctly in order to “maintain a balance between function, content and accuracy” (p. 242) while negotiating meaning in CMC environment.

Is this figure beneficial or not? If yes, How to explain that?

Additionally, from a developmental stage view point, the third research question investigated whether learners who engaged in negotiation process in CMC environment (using interactional modifications) showed a greater increase in production of developmentally more advanced questions than learners who participated in face-to-face interaction (without interactional modifications). As illustrated in figure 2, the overall frequency of all the questions in both environments revealed that CMC environment generated all stages of questions: stage 4 questions were generated most frequently (58%), followed by stage 5 questions (28%), stage 3 questions (6%), stage 6 questions (6%) and finally stage 2 questions (2%). However, Class environment generated just 2 stages of questions: stage 4 questions (67%) were generated most frequently followed by stage 5 questions (33%).

Figure 2. Percentage of questions in different stages in the whole data

Explanation why such a thing happened? By considering the fact that the same communicative tasks were administered to both group, CMC environment encourage the production of more questions than Class environment. Not surprisingly, the CMC medium-specific characteristics such as the absence of nonverbal cues together with the absence of time limitation imposed on participants may justify the difference in the production of questions in these two groups.

The table above shows that the majority of students in both CMC and classroom environment were in stage 5. In CMC environment, of the 4 students who were in stage 4 during the pretest, 3(17%) progressed to stage 5; and of 13 students who were in stage 5, 2(11%) went up to stage 6, 9(50%) stayed in stage 5 and 2(11%) went down one stage during posttest. However, in Class environment of 13 students who were in stage 5 during the pretest, no students progressed to stage 6 although 12 (67%) remained in stage 5 during the posttest. During delayed posttest, in CMC environment, totally 9 (50% students) progressed: 4(22%) from stage 4 to stage 5 and 4(22%) from stage 5 to stage 6 and 1(6%) from stage 2 to stage 4, although 9(50 %) stayed in stage 5. Nevertheless in Classroom environment, only 3(17%) students made progress from stage 4 to stage 5 and majority of the participants (67%) stayed in the same stage assigned during the pretest: 3(17%) students stayed in stage 4 together with 9(50%) students stayed in stage 5. According to the data from this study, it seemed that CMC environment outperformed the Classroom environment in generating developmentally more advanced questions.

In addition by considering “Question development” as stage increase on either posttest or delayed posttest ( as operationalized in this study), a Chi-square test for independence was conducted (See Table 7), in order to determine whether there was a significant relationship between group membership (CMC versus Class environment) and development in question formation.

Chi-square results (Continuity Correction=4.11, P<.05) showed that CMC environment generated a number of developmentally more advanced questions than the number in class environment. In addition, the phi coefficient (=.39) showed that the association between the group membership and improvement in question formation was fairly strong.

Therefore the results of this study indicated that the learners who participated in CMC environment and used a variety of interactional modifications produced a number of questions that were developmentally more advanced than the questions produced by participants in Classroom environment without the opportunity for generating interactional modifications.

Conclusion

Analysis of the chat transcripts indicated that Iranian EFL learners did engage in negotiation process through SCMC. Using this technology, they were able to carry out a verity of communicative tasks cooperatively in the target language outside the classroom. Moreover confronted by communication problems, Iranian learners employed a variety of interactional modifications to solve them. Further analysis revealed that under special conditions online negotiation may enhance accuracy.

Considering all the facts it would be inconceivable for CMC environment to replace the typical language classroom. And yet, it is also an undeniable fact that CMC technology can provide a very enjoyable and fruitful learning environment for EFL learners. Since like face-to-face interaction, CMC technology could provide an invaluable interactive learning environment, it might offer a solution to some of the greatest challenges faced by foreign language education such as the lack of sufficient practice time in the classroom and lack of exposure to the target language outside of the classroom. However, much more research is needed in order to shed light on the significant contribution of CMC technology to the foreign language education.