\nAllameh Tabataba\u2019i University, Iran
\n<zandmoghadam
![\"atmark\"](\"http:\/\/www.tesl-ej.org\/atmark.png\")
atu.ac.ir><\/p>\nFatemeh Mohandes Samani<\/strong> Although the effective role of instruction in developing L2 learners\u2019 pragmatic competence has been acknowledged in the literature, very few studies have investigated the effect of a specific language teaching method on the development of pragmatic competence. Thus, to address the existing gap, this study was conducted (1) to investigate the effectiveness of task-based instruction (TBI) on the development of pragmatic competence and (2) to see the effect(s) of information-gap, reasoning-gap, and opinion-gap tasks on the development of EFL learners\u2019 pragmatic production, metapragmatic awareness, and comprehension of implicature. To this end, 60 homogeneous intermediate EFL learners were selected and assigned to three groups randomly. The twenty members of each group received instruction on the three speech acts of request, refusal, and apology through one task type, i.e., in group #1, through information-gap, in group #2, through opinion-gap, and in group #3 through reasoning-gap tasks. The findings confirmed the positive effect of TBI on EFL learner\u2019s pragmatic competence. Also, learners in the information-gap task group outperformed the other groups on pragmatic production and metapragmatic awareness, but there was no significant difference among the groups regarding their comprehension of implicature. This study has implications for English language instructors, materials developers, and researchers.<\/p>\n Keywords<\/strong>: Task, Task-based Instruction, Pragmatic Competence, Pragmatic Production, Metapragmatic Awareness, Implicature<\/p><\/blockquote>\n In foreign language classrooms, usually the focus of instruction is on explaining grammar rules, vocabulary items, and some related exercises; consequently, the communicative aspects of language are overlooked (Derakhshan, 2014). Knowing the rules and a set of words may enable learners to create a grammatically correct sentence; however, this is not the aim of language learning. Learners also need to know how language is used appropriately in the given context (Bardovi-Harlig, 2012); this is the ability that is known as pragmatic competence. However, instructors often neglect to teach about pragmatic features of the second language (L2) (e.g., Derakhshan & Arabmofrad, 2018; Derakhshan & Eslami\u2010Rasekh, 2015; Morrow, 1996; Olshtain & Cohen, 1990; Vellenga, 2011). Moreover, whereas teachers pay attention to lexical and grammatical errors in an L2 learner\u2019s speech or writing, inappropriate pragmatic performance is usually not within the scope of what teachers focus on (Schauer, 2006). Therefore, there is a range of proficient learners who know many vocabulary items and grammar rules and make well-ordered sentences out of them; however, their final output is not contextually and situationally appropriate (Bardovi-Harlig et al., 1990). According to Bardovi-Harlig (1999, p. 686), \u201chigh levels of grammatical competence do not guarantee concomitant high levels of pragmatic competence.\u201d This inappropriacy is due to the lack of or deficiency in L2 learners\u2019 pragmatic competence. The rationale behind this phenomenon is that despite the positive effect of instruction in developing second language pragmatic competence (Alc\u00f3n-Soler, 2015; Bardovi-Harlig, 2013; Jeon & Kaya, 2006; Rose, 2005; Taguchi, 2009; Takimoto, 2006; Yoshimi & Wang, 2007), the instruction of pragmatics is insufficient or even pragmatic features are not taught at all in L2 classes (Bardovi-Harlig et al., 1991; Billmyer, 1990; Derakhshan & Eslami-Rasekh, 2020; Derakhshan & Shakki, 2020; Olshtain & Cohen 1990; Soler & Martinez-Flor, 2008).<\/p>\n Rather than just acknowledging this deficiency in the L2 instructional program, it would appear to be a propitious moment for language educators to develop a comprehensive and well-structured methodology in order to better foster pragmatic competence among L2 learners. A significant number of studies show that both explicit and implicit instruction of pragmatic features can be effective (e.g., Alc\u00f3n-Soler, 2012, 2015; Alc\u00f3n-Soler & Guzzman, 2010; Derakhshan, 2014; Derakhshan & Eslami-Rasekh, 2015; Derakhshan & Shakki, 2021; Eslami-Rasekh & Liu, 2013, Ghobadi & Fahim, 2009; House, 1996; Kubota, 1995; Shakki et al., 2020; Takahashi, 2001; Tateyama et al., 1997) though explicit instruction has been observed to be more effective in many pedagogical interventions (Plonsky & Zhuang, 2019). Besides, since different methods of instruction can influence pragmatic competence differently (Billmyer, 1990; Eslami-Rasekh et al., 2004; Rose & Ng, 2001; Tateyama, 2001; Wishnoff, 2000), introducing new, genuine, and innovative methods \u2013 which involve communicative approaches and consider the learners\u2019 role in the process of instruction \u2013 can pave the way for pragmatics instruction in an EFL\/ESL context.<\/p>\n One approach to communicative language teaching which has been found to be highly productive in helping to develop learners\u2019 communicative competence is task-based language teaching (TBLT). In this approach, the learner acquires the language through \u201cusing the language in carefully structured situations\u201d (Nunan, 2014, p. 460). Also, negotiation of meaning, which results in attention to form in doing tasks, has been shown to lead to the development of L2 acquisition (Ellis, 2003). Despite the proven benefits of TBLT, there have not been many studies investigating the effect of tasks and task-based instruction (TBI) on the development of pragmatic production and comprehension (Derakhshan, 2014; Tajeddin et al., 2012; Takimoto, 2012). In addition, the possible effect of TBI on the development of metapragmatic awareness and the comprehension of implicature among L2 learners appears not to have been investigated.<\/p>\n Thus, the main objective of the present study was to examine the effect of TBI on the development of pragmatic competence among EFL learners. To do so, the effects of three types of tasks, namely information-gap, reasoning-gap, and opinion-gap on the development of the speech acts of request, refusal, and apology among Iranian EFL learners were investigated. Also, the effects of these three different types of tasks on EFL learners\u2019 metapragmatic awareness and the comprehension of implicature were examined. Therefore, the following research questions were formulated:<\/p>\n How do EFL learners given either information-gap, reasoning-gap, or opinion-gap tasks compare with regard to:<\/p>\n Pragmatic competence has been studied from different perspectives and many aspects like pragmatics development (Bergen & Grodner, 2012; Kasper & Rose, 2002), pragmatic comprehension and production (Bouton, 1994; Derakhshan & Arabmofrad, 2018; Derakhshan & Eslami-Rasekh, 2020; Kubota, 1995; Takahashi, 2001), pragmatics across cultures and languages (Takahashi & Beebe, 1987, Zand-Moghadam & Adeh, 2020), pragmatics strategies (Cohen, 2005, 2019; Derakhshan et al., 2021; Malmir & Derakhshan, 2020; Tajeddin & Malmir, 2015), and the explicit or implicit instruction of pragmatics (Derakhshan & Shakki, 2020; Fukuya & Zhang, 2002; House, 1996; Rose & Ng, 2001; Takahashi, 2001; Tateyama, 2001), have been investigated in the literature. Most of these studies have investigated the effect of intervention on the development of different dimensions of pragmatic competence among L2 learners. In 2015, Taguchi found that in 58 studies, the effective role of L2 pragmatics instruction is confirmed. In a meta-analysis, Yousefi and Nassaji (2019) analyzed 39 studies and concluded that instruction improves L2 learners\u2019 prgamtic production and comprehension. Many of the meta-analyses (e.g., Badjadi, 2016; Derakhshan & Shakki, 2021; Plonsky & Zhuang, 2019) also indicated that not only L2 pragmatics instruction is beneficial, but also the explicit teaching of L2 pragmatic features is more effective. For instance, in the most recent meta-analysis on 17 studies, Derakhshan and Shakki (2021) found that there is an overall large effect size on the effectiveness of the instruction of request (g = 1.48) in an Iranian context. They also reported that some variables such as gender and treatment type were found to be a moderator for this effectiveness. Moreover, it was found that the male group produced a larger effect size (g = 3.09) than the female one (g = 1.10), and regarding treatment types, the explicit group yielded a larger effect size (g = 1.53) than the implicit one (g = 1.20).<\/p>\n Eslami-Rasekh et al. (2004) investigated the effect of explicit instruction on advanced EFL learners\u2019 speech act comprehension. The results revealed that the learners\u2019 speech act comprehension improved significantly. In another study, Eslami-Rasekh and Eslami-Rasekh (2008) investigated the effectiveness of metapragmatic instruction on the speech act of apology. To do so, they randomly assigned 52 advanced EFL learners to an experimental and a control group. The researchers utilized an eight-item DCT and an error recognition test (ERT) as pretests and posttests. The results showed the improvement of the experimental group in the performance and recognition of the speech act of apology. Salehi (2011) conducted a study on 40 university students to investigate the effect of explicit and implicit instruction on the acquisition and performance strategies for the speech acts of apology and request. The results revealed pragmatic gains in both groups, but not at a level of statistical significance. A study was conducted by Nguyen et al. (2012) to compare the explicit and implicit modes of instruction, focusing on criticism. It was observed that both groups\u2019 pragmatic performance, compared with a control group, improved; however, the explicit group could gain more than the implicit group. However, in 2012, Li focused on the speech act of request while implementing CALL and observed that the implicit group could outperform the other groups.<\/p>\n Mart\u00ednez-Flor and Alc\u00f3n-Soler (2008) studied the effects of instruction on developing pragmatic awareness among EFL learners. They studied 81 university students in Spain and observed that instruction could develop learners\u2019 pragmatic awareness. Alc\u00f3n-Soler and Guzm\u00e1n (2010) also tried to investigate whether instruction can develop learners\u2019 pragmatic awareness. Ninety-two graduates in translation participated in their study. The speech act of refusal was chosen as the pragmatic component to observe, and verbal reports were the means to analyze the ability of the participants in comprehending and producing refusals. The results revealed that the differences in attention before and after instruction were significant and that pedagogical intervention was helpful. Alc\u00f3n-Soler and Guzman Pitarch (2010) claimed that instruction can develop awareness among language learners.<\/p>\n Conversational implicature was introduced by Grice in 1975 to refer to the implied meaning that the addressee infers based on the contextual clues. It is \u201c \u2026 the inference a hearer makes about a speaker\u2019s intended meaning that arises from their use of the literal meaning of what the speaker said, the conversational principle and its maxims (Paltridge, 2006, p. 70).\u201d According to Taguchi and Yamaguchi (2019), implicature studies are of three types. In the first type, only Grice\u2019s maxims are taken into consideration (e.g., Bouton, 1992, 1994; Roever et al., 2014). In the second type, \u201cthe comprehension of direct and indirect speech acts\u201d (Taguchi & Yamaguchi, 2019, p. 32) is studied (e.g., Cook & Liddicoat, 2002; Garcia, 2004). The last type of studies investigated irony (e.g., Yamanaka, 2003). As far as instruction of pragmatics is concerned, not many studies have been conducted on conversational implicature (e.g., Bouton, 1999; Derakhshan, 2014, 2019; Desilla, 2012, Kubota, 1995; Lee, 2000; Padilla Cruz, 2013; Taguchi, 2007), the results of which support the positive effect of instruction on implicature development. Bouton (1999) observed the positive effect of explicit teaching of implicature to language learners. In a study, Kubota (1995) observed that explicit instruction and the use of consciousness-raising tasks would have a positive effect on Japanese university students\u2019 understanding of conversational implicature.<\/p>\n A study was conducted by Mitchelson (2011) to investigate the comprehension of implicature in an L2 setting. In doing so, two groups of nineteen native and non-native speakers of English participated in the study. They were provided with five dialogues that contained the flouting of Grice\u2019s four maxims. The participants had to answer some open-ended questions. While native speakers outperformed non-native participants in comprehending implicature and specifying the flouts of Grice\u2019s maxims, non-native speakers showed an interest in implicature skill development. They were the potential for improvement throughout instruction.<\/p>\n In the context of Iran, Mirzaei et al. (2016) studied 90 EFL undergraduate Iranian students and compared the three groups of computer-mediated communication (CMC), social media networks (SMNs), and a control group in terms of implicature comprehension. Their findings revealed that treatments of both types could improve the participants\u2019 ability to comprehend conversational implicatures. The effect of some methods of pragmatic instruction, such as metapragmatic consciousness-raising, translation, etc., on 51 EFL learners\u2019 comprehension of implicature was also studied by Derakhshan and Eslami-Rasekh (2020). They observed that instruction of any type could have a positive effect on the participants\u2019 ability to comprehend implicatures. It is also found out that implicature can be influenced by a couple of factors, such as \u201ccultural background, conventionality, the degree of formulae in implicatures, L2 learners\u2019 length of exposure to the target context, and L2 learners\u2019 general proficiency in the target language\u201d (Mirzaei et al., 2016, p. 146).<\/p>\n Although task-based instruction has been studied by many well-known researchers since the late 1990s, very few studies have focused on the effect of this communicative approach on the development of pragmatic competence among EFL or ESL learners. As Mart\u00edn -Laguna (2020, p. 28) believes, \u201cresearch on the intersection between TBLT, pragmatics, and multilingualism is still in its initial stages.\u201d<\/p>\n Investigating a task-based tutorial program, Winke and Teng (2010) found that the pragmatic competence of 19 learners of Chinese developed in eight weeks. In a pioneering study, Tajeddin et al. (2012), examined the effect of TBLT on 75 Iranian university students\u2019 pragmatic production and metapragmatic awareness. The treatment was given in the pre-task, while-task, and post-task phases. A written discourse completion test and a metapragmatic awareness questionnaire were also used to collect data. The results indicated that the use of tasks in the framework of TBLT was effective in developing the participants\u2019 pragmatic production and metapragmatic awareness. The effect of TBI on implicatures, pragmatic routines, and speech acts was investigated by Shoushinasab (2013). The study was conducted on 100 intermediate EFL students (50 male & 50 female) at a language institute in Ahvaz, Iran. The researcher concluded that TBI, though effective in general, did not yield the same results as concerns pragmatic competence, i.e., different pragmatic components were developed differently. The effect of web-based tasks on study-abroad students\u2019 pragmatic comprehension was studied by Teng and Fei in 2013. It was observed that web-based tasks were effective in developing their knowledge of L2 pragmatics, especially speech acts.<\/p>\n The production of pragmatic markers in relation to task types was studied by Neary-Sundquist in 2013. She observed that there is a relationship between proficiency level and the use of pragmatic markers. Pragmatic-focused tasks were also studied by Mart\u00edn -Laguna (2014). She observed that there is more attention to pragmatics in the interaction between the teacher and the learners, rather than between the learners themselves. Tajeddin and Hosseinpur (2014) also conducted a study to see the possible impact of different consciousness-raising tasks on the acquisition of the speech act of request by EFL learners. The main instrument was a written discourse completion test. The results revealed that instruction in any mode has positive effects on developing students\u2019 speech act production. Taguchi and Kim (2016) explored the impact of dialog construction tasks on 74 Korean learners of English, focusing on the speech act of request. The researchers observed the positive effect of collaborative dialog in developing request head acts in the collaborative group.<\/p>\n By reviewing the literature, it was made clear that only a few studies have investigated the effect of task-based instruction and different task types on the development of L2 learners\u2019 pragmatic competence and that the main components of pragmatic competence have not been taken into consideration in the previous studies. Thus, the present study not only investigated the effect of task-based instruction and the main types of tasks on Iranian EFL learners, but also considered speech act production, metapragmatic knowledge and the comprehension of implicature as the main building blocks of pragmatic competence.<\/p>\n The sample in the present study included 60 (33 female and 27 male) intermediate English students who were selected based on their proficiency level from a private language institute in Tehran, Iran. Their level of proficiency had already been determined by the institute. Their ages ranged from 16 to 46. These 60 participants were randomly assigned to one of the three treatment groups, each receiving a different task: information-gap, opinion-gap, or reasoning-gap task. There were 20 participants in each group. In that institute, the American English File <\/em>series (Latham-Koenig<\/a> et al., 2013) was being used as the main textbook series. These books come in 6 volumes (from Starter to Book 5), and the participants of the present study read Book 3.<\/p>\n Three different instruments were used in this study as both the pretest and the posttest.<\/p>\n Written discourse completion task (WDCT).<\/strong> The written discourse completion task (WDCT) used in the present study was constructed by Tajeddin et al. (2012). It consisted of 15 items, including five situations for each of the speech acts of request, refusal, and apology. The situations covered various topics and checked for control over the following factors influencing pragmatics: social status, degree of familiarity, and the degree of imposition\/severity. The participants had one minute to provide their answers for each situation.<\/p>\n Metapragmatic awareness questionnaire (MPAQ).<\/strong> In the present study, the MPAQ developed by Jianda (2006), which was confirmed to have a sufficient degree of reliability and validity, was utilized in the same situations as in the WDCT, including 15 items; five items for each of the speech acts of request, refusal, and apology. The participants had to read the situations and answer three questions. The first question was on the degree of severity or imposition, the second was about the status of the interlocutors, and the third about the degree of social distance or familiarity between the interlocutors. The learners had to answer the first and the third questions based on a five-point Likert scale, and choose the correct answer from among three choices for the second question.<\/p>\n Implicature comprehension test (ICT).<\/strong> The ICT was constructed by Derakhshan (2014) and consisted of 26 multiple-choice items. It was used as a test of how well learners perceived what was implied pragmatically in each situation. Each item included a short conversation and a description of the situation in which the conversation took place, followed by a question. For every item, the participants were required to listen to a conversation and choose the correct choice based on their comprehension of the conversation.<\/p>\n Instructional materials.<\/strong> In this study, instruction was given through the use of information-gap, reasoning-gap, and opinion-gap tasks. The rationale behind selecting these three types of tasks was that they were considered cognitive tasks in Ellis\u2019s (2003) model. Information-gap tasks involve \u201ca transfer of given information from one person to another or from one form to another or from one place to another\u201d (Ellis, 2003, p. 213). Realization of such tasks became possible through an information transfer activity (Widdowson, 1978; as cited in Ellis, 2003) and standard information-gap activity. Reasoning-gap tasks, according to Prabhu (1987, p. 46), involve \u201cderiving some new information from given information through the processes of inference, deduction, practical reasoning, or a perception of relationships or patterns.\u201d Finally, opinion-gap tasks involve \u201cidentifying and articulating a personal preference, feeling, or attitude in response to a given situation\u201d (Ellis, 2003, p. 213). These tasks could engage the students in collaborative cooperation as well as individual performances.<\/p>\n To design the instructional tasks, the researchers used the available examples and samples, textbooks for conversational courses, and some websites for English language teaching (e.g., Let\u2019s Talk, Books One & Two<\/em>, englishforeveryonr.com). The tasks\u2019 pragmatic foci were on the three speech acts of request, refusal, and apology as they manifested themselves in listening and reading, as well as in speaking and writing. Regarding the types of tasks, the researchers designed a variety of tasks, like dialogs, role-plays, listing, classifying, discussion, etc., and made use of them in the form of a three-chapter pamphlet. To design the tasks, Ellis\u2019s (2003) model was adopted. Thus, many features like task input, information configuration, interactant relationship, interaction requirement, orientation of tasks, discourse mode and domain, etc. were taken into consideration.<\/p>\n As was mentioned above, the sample of this study was 60 intermediate EFL learners whose English proficiency level had already been determined by a language institute in Tehran. These participants were randomly assigned to three groups of information-gap task (IGT), reasoning-gap task (RGT), and opinion-gap task (OGT). Each session in that institute took two hours and 15 minutes of instruction, but, every session, only 45 minutes to one hour was dedicated to teaching the designed tasks. The data collection procedure for the present study lasted for 11 sessions. In the first session, the WDCT, MPAQ, and ICT were administered before the treatment. From the second session up to the end of the tenth session, each group received instruction on the three speech acts through information-gap tasks, reasoning-gap tasks, and opinion-gap tasks, respectively. Three sessions were required for each of the speech acts of refusal, request, and apology to be instructed, and, on the whole, instruction of the speech acts lasted for nine sessions. In the eleventh session, the WDCT, MPAQ, and ICT were administered as the posttest.<\/p>\n In this study, the findings from the pretests and the posttests and also the students\u2019 performance in each group were compared and contrasted with each other to provide reasonable answers to the questions. To address the first research question, first, the participants\u2019 productions of the speech acts of request, refusal and apology in the pre-and posttest (WDCTs) were rated based on a 6-point Likert scale (from 0 to 5) developed by Taguchi (2006). Second, each group\u2019s performance on a WDCT in the pretest was compared with its performance on the same WDCT in the posttest using a multivariate ANOVA (MANOVA). Finally, the three groups\u2019 pragmatic productions were compared with each other.<\/p>\n To address the second research question, the participants of the present study had to answer the Metapragmatic Awareness Questionnaire (MPAQ), which was developed by Jianda (2006). The participants were asked to read the situations and answer the three items. The first item was on the degree of severity or imposition of the situation; the second item was about the interlocutors\u2019 status, and the third item investigated the degree of social distance or familiarity between the interlocutors. Learners had to give their answers to the first and the third item, based on a five-point Likert scale. The second item was in the form of a multiple-choice item, and the learners were supposed to choose the correct answer from among the three options. Thus, in order to get precise results, each of the three items was analyzed separately. To investigate whether the three groups were homogenous in terms of their metapragmatic awareness before the main study, a multivariate ANOVA (MANOVA) was run. After ensuring the homogeneity of the groups in the pre-test, a multivariate ANOVA (MANOVA) was run to compare the three groups\u2019 means on the posttest.<\/p>\n To address the third research question, the participants\u2019 responses to the Implicature Comprehension Test (ICT) at the beginning and the end of the study were compared through a one-way ANOVA. Before conducting the test, the homogeneity of the participants was also ensured.<\/p>\n This first question compared the pragmatic performance of the IGT, RGT, and OGT groups on the request, refusal, and apology speech acts. The results of the paired-samples t-test (t<\/em> (19) = 17.31, p <\/em>= 0.000, r <\/em>= 0.97, representing a large effect size) indicated that the IGT group had a significantly higher mean on the posttest of total WDCT (M <\/em>= 64.50, SD <\/em>= 3.44) than the pretest (M <\/em>= 52.10, SD <\/em>= 3.50). The OGT group\u2019s answers to the WDCTs in the pre-and posttest were also scored and compared. The results of the paired-samples t-test (t<\/em> (19) = 15.01, p <\/em>= 0.000, r <\/em>= 0.96, representing a large effect size) indicated that the OGT group had a significantly higher mean on the posttest of total WDCT (M <\/em>= 61.75, SD <\/em>= 3.30) than the pretest (M <\/em>= 51.40, SD <\/em>= 3.56). As for the RGT group, their answers to the WDCTs in the pre- and posttest were scored and then compared to investigate the possible difference before and after using such tasks. The results of the paired-samples t-test (t<\/em> (19) = 12.86, p <\/em>= 0.000, r <\/em>= 0.947, representing a large effect size) indicated that the RGT group had a significantly higher mean on the posttest of the total WDCT (M <\/em>= 61.75, SD <\/em>= 4.82) than the pretest (M <\/em>= 52.05, SD <\/em>= 4.58) (Table 1).<\/p>\n Table 1. Comparison of the Three Groups (Pragmatic Production; Pretest vs. Posttest).<\/strong><\/p>\n IGT<\/td>\n OGT<\/td>\n RGT<\/td>\n Then, before running the MANOVA, the homogeneity of covariance was ensured through the Box\u2019s test (Box\u2019s M <\/em>= 8.70, p<\/em> = 0.78). The non-significant result of Levene\u2019s tests, i.e., the second assumption, also indicated that the groups enjoyed homogenous variances on the pretests of the speech acts of request (F <\/em>= 0.16, p <\/em>= 0.84), refusal (F <\/em>= .86, p <\/em>= 0.42) and apology (F <\/em>= 0.63, p <\/em>= 0.53). Therefore, MANOVA was run.<\/p>\n The results of MANOVA (F<\/em> (6, 112) = 1.39, p <\/em>= 0.22, Partial \u03b7 <\/em>= 0.07, representing a moderate effect size) indicated that there were not any significant differences among the three groups\u2019 means on the pretests of the speech acts of request, refusal, and apology. Thus, the next comparison was made among the three groups in the posttest. The results indicated that (a) there were not any significant differences among the means of the IGT (M <\/em>= 21.60), OGT (M <\/em>= 20.65) and RGT (M <\/em>= 21.25) groups\u2019 means on the posttest of the speech act of request (F<\/em> (2, 57) = 2.41, p <\/em>= 0.099, Partial \u03b7 <\/em>= 0.07, representing a moderate effect size); (b) there were not any significant differences between the means of the IGT (M <\/em>= 20.65), OGT (M <\/em>= 19.60) and RGT (M <\/em>= 19.60) groups\u2019 means on the posttest of the speech act of refusal (F<\/em> (2, 57) = 2.03, p <\/em>= 0.14, Partial \u03b7 <\/em>= 0.067, representing a moderate effect size), and (c) there were significant differences between the means of the IGT (M <\/em>= 22.25), OGT (M <\/em>= 21.45) and RGT (M <\/em>= 20.95) groups\u2019 means on the posttest of the speech act of apology (F<\/em> (2, 57) = 3.78, p <\/em>= 0.029, Partial \u03b7 <\/em>= 0.11, representing a moderate to large effect size) (Table 2). To recognize where the difference exists, a post-hoc Scheffe\u2019s test was also used. The results of the post-hoc Scheffe\u2019s test indicated that from among the pairwise comparisons done, there was only one significant difference between the IGT and RGT groups\u2019 means on the posttest of apology; the IGT group (M <\/em>= 22.25) had a significantly higher mean than the RGT group (M <\/em>= 20.95) on the posttest of the speech act of apology (MD <\/em>= 1.30, p <\/em>= 0.030).<\/p>\n Table 2. Comparison of the Three Groups (Pragmatic Production in Posttests).<\/strong><\/p>\n The second question looked at the impact of task group (IGT, RGT, or OGT) on the learners\u2019 metapragmatic awareness. To address this question, first, it was observed that at the beginning, the three groups were homogeneous: Box\u2019s test (Box\u2019s M <\/em>= 8.48, p <\/em>= 0.79) and Levene\u2019s test [request (F <\/em>= 0.62, p <\/em>= 0.53), refusal (F<\/em>=.03, p<\/em>=.96), and apology (F <\/em>= 3.11, p <\/em>= .05)]. Then, the results of MANOVA (F<\/em> (6, 112) = 1.12, p <\/em>= 0.354, Partial \u03b7 <\/em>= 0.05, representing a weak effect size) indicated that there were not any significant differences between the three groups\u2019 means on the pretests of metapragmatic awareness of request (F<\/em> (2, 57) = 1.82, p <\/em>= 0.140, Partial \u03b7 <\/em>= 0.06, representing a moderate effect size), refusal (F<\/em> (2, 57) = 0.10, p = 0.900, Partial \u03b7 <\/em>= 0.004, representing a weak effect size), and apology (F<\/em> (2, 57) = 1.24, p = 0.29, Partial \u03b7<\/em> = .042, representing a weak effect size).<\/p>\n Next, the non-significant results of the Box\u2019s test (Box\u2019s M<\/em>=4.29, p=0.984) indicated that the assumption of homogeneity of covariance matrices was met. It was also observed that the groups enjoyed homogenous variances on the posttests of metapragmatic awareness of request (F<\/em> = 2.61, p <\/em>= 0.082), refusal (F <\/em>= 0.54, p <\/em>= 0.584), and apology (F <\/em>= 0.19, p <\/em>= 0.827). Then, a MANOVA was run to compare the three groups on the post-test. The results (F<\/em> (6, 112) = 3.97, p <\/em>= .001, Partial \u03b7 <\/em>= 0.17, representing a large effect size) indicated that there were significant differences between the three groups\u2019 means on the posttests of metapragmatic awareness of request, refusal, and apology.<\/p>\n It can be stated that (a) there were significant differences between the means of the IGT (M <\/em>= 39.90), OGT (M <\/em>= 36.20), and RGT (M <\/em>= 36.75) groups\u2019 on the posttest of metapragmatic awareness of request (F<\/em> (2, 57) = 6.75, p <\/em>= 0.002, Partial \u03b7 <\/em>= 0.19, representing a large effect size). The results of the post-hoc Scheffe\u2019s tests also indicated that, first, the IGT group (M <\/em>= 39.90) significantly outperformed the OGT (M <\/em>= 36.20) on the posttest of metapragmatic awareness of request (MD <\/em>= 3.70, p <\/em>= 0.005); second, the IGT group (M <\/em>= 39.90) significantly outperformed the RGT (M <\/em>= 36.75) on the posttest of metapragmatic awareness of request (MD <\/em>= 3.15, p <\/em>= 0.020). Finally, there was not any significant difference between the RGT (M <\/em>= 36.75) and OGT (M <\/em>= 36.20) groups\u2019 means on the posttest of metapragmatic awareness of request (MD <\/em>= 0.55, p <\/em>= 0.880); (b) there were significant differences between the means of the IGT (M <\/em>= 43.85), OGT (M <\/em>= 42) and RGT (M <\/em>= 41.60) groups\u2019 means on the posttest of metapragmatic awareness of refusal (F<\/em> (2, 57) = 4.84, p <\/em>= 0.011, Partial \u03b7 <\/em>= 0.145, representing a large effect size). Accordingly, it can be mentioned that, first, there was not any significant difference between IGT (M <\/em>= 43.85) and OGT (M <\/em>= 42) groups\u2019 means on the posttest of metapragmatic awareness of refusal (MD <\/em>= 1.85, p<\/em> = 0.060); second, the IGT group (M <\/em>= 43.85) significantly outperformed the RGT group (M <\/em>= 41.65) on the posttest of metapragmatic awareness of refusal (MD <\/em>= 2.20, p<\/em> = 0.020). Finally, there was not any significant difference between the RGT (M <\/em>= 41.65) and OGT (M <\/em>= 42) groups\u2019 means on the posttest of metapragmatic awareness of refusal (MD <\/em>= 0.35, p <\/em>= 0.900), and (c) there were significant differences between the means of the IGT (M <\/em>= 36.80), OGT (M <\/em>= 35.65) and RGT (M <\/em>= 34.50) groups\u2019 means on the posttest of metapragmatic awareness of apology (F<\/em> (2, 57) = 3.53, p <\/em>= 0.036, Partial \u03b7 <\/em>= 0.11, representing a moderate to large effect size). Accordingly, it can be claimed that first, there was not any significant difference between the IGT (M <\/em>= 36.80) and OGT (M <\/em>= 35.65) groups\u2019 means on the posttest of metapragmatic awareness of apology (MD <\/em>= 1.15, p<\/em> = 0.419); second, the IGT group (M <\/em>= 36.80) significantly outperformed the RGT group (M <\/em>= 34.50) on the posttest of metapragmatic awareness of apology (MD <\/em>= 2.30, p <\/em>= .036). Finally, there was not any significant difference between the RGT (M <\/em>= 34.50) and OGT (M <\/em>= 35.65) groups\u2019 means on the posttest of metapragmatic awareness of apology (MD <\/em>= 1.15, p <\/em>= 0.419) (Table 3).<\/p>\n Table 3. Comparison of the Three Groups (Metapragmatics in Posttests).<\/strong><\/p>\n
\nAllameh Tabataba\u2019i University, Iran
\n<samani_f91yahoo.com><\/p>\n\n
Abstract<\/h3>\n
\n
Literature Review<\/h3>\n
Pragmatic Production and Metapragmatic Awareness<\/h4>\n
Implicature<\/h4>\n
Task, Task-based Instruction, and Pragmatic Competence<\/h4>\n
Method<\/h3>\n
Participants and Research Setting<\/h4>\n
Instruments<\/h4>\n
Data Collection Procedure<\/h4>\n
Data Analysis<\/h4>\n
Results<\/h4>\n
The Comparability of Speech Act Production across Task Groups<\/h4>\n
\n\n
\n Group<\/td>\n Test<\/td>\n Mean<\/td>\n N<\/td>\n Std. Deviation<\/td>\n Std. Error Mean<\/td>\n T<\/td>\n df<\/td>\n P<\/td>\n<\/tr>\n \n \nTotal WDCT<\/p>\n Posttest<\/td>\n 64.50<\/td>\n 20<\/td>\n 3.44<\/td>\n .76<\/td>\n 17.31<\/td>\n 19<\/td>\n .000<\/td>\n<\/tr>\n \n Pretest<\/td>\n 52.10<\/td>\n 20<\/td>\n 3.50<\/td>\n .78<\/td>\n <\/td>\n <\/td>\n <\/td>\n<\/tr>\n \n \nTotal WDCT<\/p>\n Posttest<\/td>\n 61.75<\/td>\n 20<\/td>\n 3.30<\/td>\n .73<\/td>\n 15.01<\/td>\n 19<\/td>\n .000<\/td>\n<\/tr>\n \n Pretest<\/td>\n 51.40<\/td>\n 20<\/td>\n 3.56<\/td>\n .79<\/td>\n <\/td>\n <\/td>\n <\/td>\n<\/tr>\n \n \nTotal WDCT<\/p>\n Posttest<\/td>\n 61.75<\/td>\n 20<\/td>\n 4.82<\/td>\n 1.07<\/td>\n 12.86<\/td>\n 19<\/td>\n .000<\/td>\n<\/tr>\n \n Pretest<\/td>\n 52.05<\/td>\n 20<\/td>\n 4.58<\/td>\n 1.02<\/td>\n <\/td>\n <\/td>\n <\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n \n\n
\n Source<\/td>\n Dependent
\nVariable<\/td>\nType III Sum of Squares<\/td>\n Df<\/td>\n Mean Square<\/td>\n F<\/td>\n Sig.<\/td>\n Partial Eta Squared<\/td>\n<\/tr>\n \n Group<\/td>\n Post WDCT Req<\/td>\n 9.23<\/td>\n 2<\/td>\n 4.61<\/td>\n 2.41<\/td>\n .099<\/td>\n .07<\/td>\n<\/tr>\n \n Post WDCT Ref<\/td>\n 14.70<\/td>\n 2<\/td>\n 7.35<\/td>\n 2.03<\/td>\n .140<\/td>\n .06<\/td>\n<\/tr>\n \n Post WDCT App<\/td>\n 17.20<\/td>\n 2<\/td>\n 8.60<\/td>\n 3.78<\/td>\n .029<\/td>\n .11<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n The Comparability of Learners\u2019 Metapragmatic Awareness by Task Group<\/h4>\n