INTRODUCTION

Horwitz et al. (1986) argued that language learning contexts produce a unique type of anxiety which is different from its other types. Moreover, they claim that Foreign Language Anxiety (FLA) should not be simply perceived as “fears transferred to foreign language learning” but rather conceptualized as “a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process” (Horwitz et al., 1986, p.128). It has been claimed that FLA is widely common among EFL learners (MacIntyre, 1995) and we can find it in performance of any language skills. As a result, students might not participate actively in class activities since they feel fear and insecurity (Llinás & Garau, 2009). This implies that debilitating language anxiety may produce a vicious circle; the more students avoid participation in classes to evade anxiety the less they would improve their performance, which, in turn, could produce more anxiety.

On the other hand, dialogic teaching is a pedagogical approach that allows teachers and students to interact collaboratively and build on each other’s ideas to improve learning outcomes (Hennessy et al., 2011). It mainly focuses on highlighting the dialogic processes wherein both teachers and students act as inquirers in dialogic exchanges during classroom instruction (Haneda & Wells, 2008; Lyle, 2008). On the contrary, in traditional didactic teaching, teachers tend to disseminate information while students act as passive receivers of knowledge with limited participation in classroom dialogues (Alexander, 2008; Hennessy, 2017; Lee, 2016; Skidmore, 2006).

Derived from the principles of dialogic teaching, dialogic tasks, as opposed to monologic ones, have been defined as tasks that engage both teachers and learners in the co-construction of knowledge and improvement of learning in the target language. Hence, students and teachers work collaboratively to co-construct meanings over successive utterances to achieve teaching and learning goals. This being so, students are likely to encounter varied perspectives on a topic under discussion (Haneda & Wells, 2013). As a frequently used mode of oral language in real life and pedagogic contexts, a dialogic task is “prototypically a joint enterprise involving more than one person” (Cameron, 2001, p. 87), with the speakers’ taking turns in conversation. Cameron (2001) suggested that attributes that make a dialogue distinct from a monologue include between-turn pauses, interruptions by the second speaker, and simultaneous talk.

Dialogic tasks engage students in practical and meaningful contexts in which they have to use the language interactively. Every student’s voice is heard in such activities. Thus, dialogic tasks combined with TBLT seem to be useful for removing the students’ anxiety in foreign language classrooms. However, other contextual and individual factors might alter the efficiency of these tasks. This study was an attempt to investigate the potential effect of dialogic tasks on language learning anxiety with a focus on the moderating effects of gender and proficiency levels.

LITERATURE REVIEW

Theoretical Background

Task-based approaches to language teaching, known as task-based language teaching (TBLT), began in the 1980s. It initially emerged from the constructs of communicative approaches to language teaching which were first crystallized by Brumfit and Johnson (1979). Some benefits of TBLT include promoting learners to see language as functional (Bygate et al., 2015) and improving learners’ interlanguage through hypothesis testing and negative feedback during task performance (Long, 1996; Swain & Lapkin, 1998). This being so, TBLT, which regards tasks as the core unit of analysis in syllabus design, may be considered the most researched pedagogical approach in the field of SLA (Long, 2014).

The review of the literature on TBLT shows that there have been two main approaches to task-based studies within SLA: the interactionist and the information-processing approaches. The former is rooted in the interaction hypothesis (Long, 1996), and the latter from the Cognitive Approach (Skehan, 1998) or the Cognition Hypothesis (Robinson, 2001,2007). These two approaches differ from each other regarding the focus of the investigation; however, they both interrogate learner language production.

The interactionist approach has to do with those activities that trigger the promotion of interaction, such as negotiation of meaning and feedback. This approach stems mainly from the argument that such an interaction facilitates SLA. Long (1996), in his original and then updated Interaction Hypothesis, claims that acquisition is expected to take place when learners obtain comprehensible input as a result of the opportunity to interact, especially when communication breakdown occurs. Thus, task-based studies within the interactionist perspective are characterized by interaction activities, with learners’ working in pairs or groups, and engaging in a number of different types of tasks such as making jigsaws, filling information gaps, or doing role plays, etc.

On the other hand, some researchers such as Skehan (2003), whilst supporting the key role of interaction, argue that the focus of research should extend beyond interaction. In particular, he claims that the focus of task-based studies should be on understanding the psychological processes that learners use when working on tasks (Skehan, 2003). This second approach is known as the Cognitive Approach (Skehan, 1998) or the Cognition Hypothesis (Robinson, 2001, 2007).

 Ellis (2000) provides further support in favor of this position arguing that the information-processing perspective, on which these models are based, “could be predictive and deterministic which mean that properties in a task will predispose or even induce learners to engage in certain types of language use and mental processing that are beneficial to acquisition” (p.197). This study is informed by the interactionist perspective and characterized by interaction activities such as dialogic tasks.

Nevertheless, regardless of the amount of research and data collected on TBLT, its application in L2 classroom contexts is not without challenges. It has been claimed that TBLT is an often-misunderstood concept (Carless, 2004; Ellis, 2009), because of the multiple definitions of task that exist in the literature, and the various opinions on how to appropriately run TBLT in the L2 classroom (Long, 2016; Willis & Willis, 2007). A concept closely related to the application of tasks in the language learning and teaching is anxiety.

          There are three approaches to the study of anxiety in the language learning domain including trait, state, and situation-specific perspectives (Horwitz & Young, 1991). Among these three types, situation-specific anxiety was targeted in this study. Horwitz et al. (1986) argued that language learning contexts produce a unique type of anxiety which is different from its other types. This approach examines the specific type of anxiety in a well-defined and particular situation (Horwitz, 2010) such as communication apprehension, stage fright, and role play. The literature on the topic reveals that FLA should be distinguished from other general performance and academic anxieties. Ellis (1994, p.480) defined anxiety in the language learning context as “a type of situation-specific anxiety associated with attempts to learn a second language and communicate in it”. Likewise, MacIntyre (1999, p.27) argued that language anxiety is “the worry and negative emotional reaction aroused when learning or using a second language”.

          A variety of researchers have claimed that the situation-specific anxiety approach provides a better understanding of language anxiety, compared with state and trait anxieties, since it limits the assumptions and requires participants to attribute their language anxiety to a particular source (Horwitz, 2010; Tóth, 2011). In addition, the concept of FLA, as a unique type of anxiety specific to foreign language learning, has been advocated by many SLA studies which reviewed this theory and discussed the criticisms that have been leveled against it (Trang, 2012; Zheng, 2008). Thus, this is a concept which has gained credence among FLA researchers (Cao, 2011, Fathi et al., 2021; Omidbakhsh, 2021; Ramamuruthy, 2019; Zarei & Kavyari Roustai, 2019).

Empirical Background

A quick review of the SLA literature reveals that many researchers have taken interest in examining the effects of TBLT on reducing EFL learners’ anxiety in L2 classrooms (Brennan, 2016; Mohammadi Khomjani, 2020; Omidbakhsh, 2021; Salimi & Karimi, 2019; Zarei & Mokhtari Rezaei, 2016). In a relevant study, Brennan (2016) dealt with the effects of TBLT on speaking anxiety among EFL learners. The aim of her study was twofold. First, the objective was to explore the effects of whether the L2 tasks in terms of task-type, task complexity, or sequencing of tasks had any effect on the participants’ speaking anxiety levels. The second objective was to explore whether state anxiety could be identified during task-based oral performance by using a scale designed for the current study. The results suggested that the state anxiety scales were successful at analyzing how the students felt during L2 oral task performance that had been manipulated in terms of task-type, task complexity, and sequencing. Her results have extended the research into how L2 task features, based on sequence and complexity, affect learners’ speaking anxiety levels.

         Ramamurthy (2019) focused on the effects of task-based approach on speaking anxiety among ESL students. More precisely, Ramamurthy investigated the presence of speaking anxiety among ESL learners, the relationship between language anxiety and speaking skills, and the effect of task-based approach on learners’ speaking anxiety. The findings revealed that anxiety exists among the learners in terms of communication apprehension, English language class anxiety and fear of negative evaluation. The more anxious the learners became, the lower the grades were achieved. The use of task-based approach throughout the intervention was found to have exerted positive effect on the grades achieved in the speaking posttest.

         In another very relevant study, Hidalgo (2020) analyzed to what extent the "weak" version of TBLT supplemented by sound instructional practice strategies could reduce anxiety and improve online spoken interaction among learners who were in a CEFR B2 EFL course. The length of the intervention was five weeks, and it involved twenty-nine students who took this course as a requirement to obtain their tertiary degree in Guayaquil- Ecuador. Quantitative results showed that the intervention had a large effect both on the students’ anxiety reduction, and online spoken interaction.

          In another recent study, Mohammadi Khomjani (2020) investigated the effect of a certain type of task-based activity (i.e., digital storytelling) on L2 learning motivation and reducing anxiety of Iranian EFL learners to talk to others in a foreign language. The findings of this study, based on pre-tests and post-tests in the form of questionnaires, support that TBLT-oriented activities in general and DST in particular reduce anxiety for English foreign language speakers to talk to others, but DST intervention does not affect the motivation of English foreign language speakers to talk to others.

       Moreover, Ozturk and Ozturk (2021) examined the impact of tasks coupled with mini-speeches, presentations, and scaffolded feedback on EFL learners’ foreign language speaking anxiety. Their findings showed the effectiveness of TBLT in reducing Turkish EFL learners’ anxiety. In fact, they claimed that the task-based activities resulted in around thirty percent reduction in participants’ EFL speaking anxiety.

            As it is evident, the previous studies indicated the efficiency of TBLT in a general sense, and specifically dialogic tasks not only in reducing the learner’s language learning anxiety and also speaking anxiety, but also the effectiveness of dialogic tasks in increasing learners’ participation in the classroom activities. However, nearly all these studies have ignored the moderating effects of individual and contextual factors. Hence, this study was an attempt to find out whether dialogic tasks could influence learners at different proficiency levels in the same way. Moreover, the moderating effect of gender was investigated.

PURPOSE OF THE STUDY

It can be claimed that recent research on foreign language learning anxiety in L2 classrooms has paid particular attention to the methods and techniques for reducing the level of anxiety in language classrooms. Nonetheless, it appears that more research is needed to understand the effectiveness of these techniques, taking into account a larger number of variables to gain more practical and realistic findings. To fill the existing gap in the literature, this study aimed at examining the effects of dialogic tasks on reducing Iranian EFL learners’ language learning anxiety with a focus on proficiency levels and gender as moderating individual factors. Bearing this objective in mind, this study addressed the following research questions:

1. Do dialogic tasks play a significant role to reduce Iranian EFL learners’ language learning anxiety?
2. Does proficiency levels of learners significantly moderate the probable effects of dialogic tasks on EFL learners’ language learning anxiety?
3. Does learners’ gender significantly moderate the probable effects of dialogic tasks on EFL learners’ language learning anxiety?
4. Are there any significant interaction effects between levels of proficiency and gender considering the probable effects of dialogic tasks on learners’ language learning anxiety?

METHOD

Participants

A total number of 213 male and female Iranian EFL learners within the age range of 15-19 were selected through the convenience sampling procedure out of 487 available students in three language schools located in Khafr city, Fars province, Iran. All of them were high school students and their first language was Persian. In spite of the schools’ criteria for determining the learners’ level of proficiency, the researchers ran the Oxford Quick Placement Test (OQPT) in order to ensure the proficiency levels and select the 213 participants based on the results. Consulting OQPT’s scoring criteria, the learners were divided into two main levels of proficiency (i.e., upper vs. lower intermediate) and they were assigned to the experimental and control groups. The details regarding the grouping of the participants could be seen in Table 1.

Table 1: The Number of Participants in Each Group and Subgroup

As it could be seen in Table 1, there were 106 learners in the experimental group (EG henceforth) and 107 learners in the control group (CG hereafter). The total number of lower-intermediate learners was 108, and there were 105 upper-intermediate participants in this study. In addition to these EFL learners, there were 12 teachers who taught these classes. In order to train the teacher participants for the project in the study, one of the researchers held an educational session of about two hours to clarify the detailed rules and principles of the prescribed teaching in this project. He introduced key points of dialogism and task-based instruction briefly.

Instrumentation

The following instruments were utilized in this study:

Placement Test

Oxford Quick Placement Test (OQPT) was run in order to ensure learners’ homogeneity and select 213 participants at two proficiency levels. Those learners who scored 30-39 were labeled lower-intermediate and the ones who scored from 40 to 47 were considered to be upper-intermediate. The OQPT had 60 multiple-choice questions of vocabulary, grammar, and reading comprehension. The test has been validated before and is widely used by researchers all over the globe. 

Anxiety Questionnaire

In order to determine participants’ language learning anxiety before and after the treatment, the researchers utilized an adapted version of Foreign Language Classroom Anxiety Scale (FLCAS) (Horwitz et al., 1986). The questionnaire is comprised of 31 items translated into Farsi in Likert-scale—namely 1. Strongly agree; 2. Agree; 3. Unsure; 4. Disagree; 5. Strongly disagree. It was also piloted before the commencement of the study and all the deficiencies and problems were recognized and eradicated. In order to reassure the translation reliability (internal consistency) of the questionnaire, the researchers selected 35 volunteer participants to take part in the reliability process—piloting. The results of the reliability check are depicted in Table 2 which was high enough to confirm the reliability of the test.

Table 2: Anxiety Questionnaire Reliability Statistics Cronbach’s Alpha

As can be seen, the Cronbach's alpha reliability coefficient for the questionnaire in this study was computed to be .87. As well as reliability, the content validity of the questionnaires was checked by a board of university professors teaching at Islamic Azad University, Shahreza branch, and three certified teachers in Iranian Education Department, and the necessary changes were made to remove ambiguity of some wordings.

Data Collection Procedure

At the outset of the study, the anxiety pretest was administered and the learners were asked to answer the Likert scale anxiety questionnaire in order to measure their entry amount of language learning anxiety before launching the dialogic treatment program. The pretest was held in forty-minute sessions separately for the experimental and control groups.

A week after administering the pretest, the treatment phase of the study began. In this study, the researchers utilized dialogic tasks through which students could interact meaningfully not just mechanically. This being so, sequencing picture stories were used as dialogic tasks. There were 8 sets of four picture stories which were downloaded from Twinkl website and copied to be used in the experimental and the control groups. First of all, participants in the experimental classes were categorized into groups of four. Then, each of the pictures in a set were given to one of the group members. In order to sequence the pictures correctly, the members in each group were supposed to ask and answer questions about the pictures and each student had to describe what could be seen in each picture. Once the pictures were sequenced, they were supposed to collaboratively narrate a story based on the four sequenced pictures. The treatment took 8 sessions: 2 sessions per week for a month.

In the control groups, the same set of pictures was used. However, there were no dialogic tasks. In other words, each student was supposed to sequence the pictures individually and some of them were randomly asked to narrate the stories. There were no opportunities for communication or dialogue among the classmates.

Finally, FLCAS questionnaire was administered to the participants in the experimental and control groups to measure their levels of language learning anxiety after the treatment.

DATA ANALYSIS

This study aimed at examining the effects of dialogic tasks on Iranian EFL learners' language learning anxiety with a focus on the gender and proficiency level effects. Because there are three independent variables in focus in the research questions (i.e., using dialogic tasks, proficiency level, and gender), a three-way ANCOVA was conducted to compare the anxiety posttest scores of the male and female learners in the two groups of EG and CG and at the two proficiency levels of lower- and upper-intermediate. The motive behind using ANCOVA was that it controls for any possible pre-existing differences between the groups, and then compares the (adjusted) posttest mean scores of the learners.

RESULTS

In order to answer the research questions, a three-way ANCOVA was carried out. The anxiety posttest mean scores of male and female learners at different proficiency levels in the EGs and CGs are displayed in Table 3. The total mean score for EG learners (M = 79.89) indicated a lower level of anxiety compared with that of CG (M = 89.65) learners. The total anxiety posttest mean score for the lower-intermediate learners (M = 85.53) was found to be slightly higher than the one for upper-intermediate learners. Moreover, the total mean scores for male and female learners indicated that males (M = 84.27) were a little less stressed than females (M = 85.31). The details regarding the mean scores and standard deviations of the male and female learners in the lower- and upper-intermediate subgroups of the EG and CG are also shown in the table below.

Table 3: Descriptive Statistics for Anxiety Posttest Scores of the Learners

The skewness and kurtosis values displayed in Table 3, which respectively show the skewness and peakedness of the distributions, show that the distributions of anxiety posttest mean scores for male and female learners in the lower- and upper-intermediate subgroups in EG and CG were normal, as these skewness and kurtosis values did not exceed ±2.00. After ensuring the normality assumption, the researcher made sure the assumption of homogeneity of variances was not violated (p = .13 > .05), and that the assumption of homogeneity of regression slopes was met as well. To find out whether the differences among the two groups of learners and between males and females at the two different proficiency levels were statistically significant or not, the researcher had to examine the p values in front of Groups, Proficiency, and Gender (as well as the other rows which show the interaction effects) under the Sig. column in the three-way ANCOVA table below (Table 4).

Table 4: Three-way ANCOVA for Anxiety Posttest Scores of the Learners

Regarding the answer to the first research question, as it is shown in Table 4, there was a statistically significant difference in the anxiety posttest mean scores of the EG and CG learners because the p value under the Sig. column in front of Groups was smaller than the specified level of significance (i.e., .000 < .05). This significant impact, as presented under the Partial Eta Squared column, indicated a very large effect size (.18), based on Cohen (1988, as cited in Pallant, 2010), who interpreted effect sizes as: .01 = small, .06 = moderate, and .14 = large.

       Considering the second research question, it is worth mentioning that the p value corresponding to Proficiency was found to be smaller than the alpha level of significance (.000 < .05), which shows that the learners at the lower-intermediate level of proficiency were significantly more anxious than those in the upper-intermediate level of proficiency. With regard to the third research question, the p value across from Gender was lower than the significance level (.014 < .05), implying that females suffered significantly more than males from L2 learning anxiety. 

         Finally, regarding the interaction effects among the three independent variables of the study (i.e. fourth research question), it could be observed that all the interactions (i.e., the interaction between Groups and Proficiency, the interaction between Groups and Gender, the interaction between Proficiency and Gender, and the interaction among Groups, Proficiency, and Gender) were all found to exert statistically significant impacts on the anxiety posttest scores of the Iranian EFL learners (p < .05). In sum, it was concluded that CG learners experienced greater levels of anxiety than did the EG learners. It was also revealed that upper-intermediate learners were less concerned than their lower-intermediate counterparts. Finally, it was found that female learners (as it was especially noticed for the EG) had more anxiety levels than male learners.

DISCUSSION

The main research question of the study examined the effects of dialogic tasks on Iranian EFL learners' language learning anxiety. More specifically, within this main research question of the study, the moderating effects of the learners’ level of proficiency and their gender were also considered. To address and answer the aforementioned research questions, a three-way ANCOVA was conducted to compare the anxiety posttest scores of the male and female learners in the two groups of EG and CG and at the two proficiency levels of lower- and upper-intermediate. The major findings obtained from the data analysis showed that CG learners experienced greater levels of anxiety than did the EG learners. In other words, dialogic tasks were found to have a significant effect on reducing Iranian EFL learners' language learning anxiety. Moreover, it was revealed that upper-intermediate learners experienced lower levels of anxiety than their lower-intermediate counterparts. Finally, it was discovered that female learners in this study suffered from more anxiety levels than male learners.

The aforementioned findings of this study are in line with the findings of other researchers who have confirmed the positive effects of dialogic teaching by arguing that dialogic teaching—an effective teaching approach that allows teachers and students to interact collaboratively and actively build on each other’s ideas—can not only reduce EFL learners’ levels of anxiety but also increase classroom engagement (Alexander, 2008; Hammond & Gibbons, 2005; Haneda & Wells, 2008; Hennessy et al., 2011). The lower levels of anxiety experienced by the learners who received dialogic teaching in this study can be justified by the fact that dialogic teaching is different from traditional didactic teaching in which teachers tend to disseminate information while students act as passive receivers of knowledge with limited participation in classroom dialogues (Alexander, 2008; Chow et al., 2021; Hennessy, 2017; Lee, 2016; Skidmore, 2006). In other words, by employing dialogic tasks, students and teachers work collaboratively to co-construct meanings over successive utterances to achieve teaching and learning goals. Such strong cooperation among the learners and their peers or their teachers can also play a key role in reducing EFL learners’ levels of anxiety in classrooms (Alikhani & Kiany, 2021; Schleppegrell & Moore, 2018).

Moreover, the findings lend further support to the findings of other researchers who have claimed that TBLT in general and dialogic tasks in particular can provide a safe learning environment to engage leaners through meaning-centered activity (Ellis, 2009; Kubanyiova, 2018). In essence, the lower levels of anxiety observed in the posttest scores of the EG participants who received dialogic tasks can be related to the safe learning environment created within TBLT that can engage leaners through meaning-centered activities. Kubanyiova (2018) defined a safe speaking environment as a space that treats people as a resource that need protection and care. She further explained that a safe speaking environment independently of its linguistic feature or meaning should be in the service of students’ conversational accomplishments. This non-threatening, safe learning, language-rich environment, created in L2 speaking classes through dialogic tasks, results in learners constantly using L2 during class with lower levels of L2 language learning anxiety (Hidalgo, 2020; Ramamuruthy, 2019; Richards et al., 2001). Consequently, it can be claimed that when learning occurs in a safe supportive environment learners are more likely to succeed, they feel freer to ask questions, and they share opinions and feelings more openly, and less anxiously.

Likewise, and in line with our findings, previous literature has revealed that when TBLT is implemented properly through well-designed tasks such as dialogic tasks employed in this study and supported by sound instructional practice (Nunan, 2004), lowered anxiety among learners (Boonkit, 2010; Ellis et al., 2019; Wang, 2017) and improved spoken interaction will result (Ellis et al., 2019; Knight & Barbera, 2018). Thus, it can be claimed that L2 language learning anxiety levels tended to diminish among learners who received dialogic tasks through the treatment phase because, as suggested by other researchers, these tasks have high potentials for creating less threatening learning environments (González-Lloret & Ortega, 2014; Mohammadi Khomjani, 2020; Ziegler, 2016). In other words, dialogic tasks can create a collaborative and cooperative L2 environment within which learners can show improvements in the spoken interaction through simple, easy to use, and collaborative tasks with less anxiety levels (Tavakoli, 2016). The results supported the findings of Allen and Herron (2003), and Dewaele (2010) who state that anxiety could be less important once the language learner has reached higher levels of proficiency.

As for the obtained results on gender, the findings of this study comply with those of other researchers that have claimed that gender is one of the factors influencing the language learning process and the level of anxiety among foreign language learners (Gandhimathi, 2016). More precisely, our findings revealing that, after receiving dialogic tasks, female EFL learners experienced higher levels of L2 speaking anxiety compared to their male counterparts are confirmed by other researchers who have reported similar results (Dewaele et al., 2016; Gerencheal & Horwitz, 2016). Such findings on female learners’ higher anxiety levels can be explained by the fact that many studies have found that female students seem more anxious due to the fear of failure, teachers’ corrections and negative evaluations and unpreparedness (Aydin et al., 2017; Khalid et al., 2020). In fact, it has been found that female students seem more sensitive to being laughed at by other students and unprepared speaking in front of their peers (Gandhimathi, 2016).

          It should be noted, however, that research on gender, as one of the factors influencing the language learning process and the level of anxiety among foreign language learners, shows conflicting results. In other words, contrary to our findings and those of the aforementioned researchers, there are studies in the past literature claiming that male learners suffered from higher levels of FLA (Fariadian et al., 2014), or it has been reported by others that there was no correlation between gender and foreign language anxiety (Loo & Kitjaroonchai, 2015; Piechurska-Kuciel, 2012).

On the whole, it can be claimed that gender and proficiency levels can moderate the effectiveness of dialogic tasks in lowering the language learning anxiety. Hence, it seems too simplistic to make overgeneralized conclusions on the pedagogical efficiency of teaching and learning techniques without considering the individual factors.

CONCLUSION AND IMPLICATIONS

The present study was designed to investigate the effects of dialogic tasks on Iranian male and female EFL learners’ anxiety in speaking at two proficiency levels of lower- and upper-intermediate. As such, the potential impact of dialogic tasks on anxiety across two levels of proficiency and gender was examined. It was found that CG learners experienced greater levels of anxiety than did the EG learners; that is, dialogic tasks were found to have exerted a significant effect on reducing Iranian EFL learners' anxiety. Moreover, it was revealed that upper-intermediate learners experienced lower levels of anxiety than their lower-intermediate counterparts. Finally, it was discovered that female learners in this study suffered from more anxiety levels than male learners. The cooperation among the learners and their peers and their teachers could explain the reduction in EFL learners’ levels of anxiety in classrooms (Hennessy, 2017; Alikhani & Kiany, 2021). Furthermore, the findings showed that female learners’ higher anxiety levels can be explained by the fact that female students seem more anxious due to the fear of failure, teachers’ corrections and negative evaluations and unpreparedness (Aydin et al., 2017; Khalid et al., 2020).

          It is hoped that the findings of the present study will encourage EFL teachers to pay closer and more consummate attention to the concept of dialogic teaching. In fact, bearing in mind the beneficial impacts of providing EFL learners with dialogic tasks (Tavakoli, 2016) in speaking classrooms, EFL teachers are expected to invest more time in such activities. As demonstrated by previous studies in the literature (Tavakoli et al., 2016), dialogic instruction remains a marginal component of the TL instruction, receiving only negligible attention by most EFL teachers. This being so, EFL teachers are expected to take the findings reported by experimental studies like this into account and integrate dialogic tasks into their lesson plans more than before. Moreover, it has been observed that one major difference between EFL learners and learners in a second language context is that the former cannot benefit from rich exposure to the target language and sufficient opportunities to use it for real-life purposes. Accordingly, EFL learners need to play a more active role in this regard and benefit from the activities, techniques, and the instructional approaches employed in studies like this to improve their speaking skills.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID