. Introduction

              Second language speaking is a priority for many L2 or foreign-language learners. The reason lies in the fact that speaking, as a significant component of the target language, is the prime means of communication (Hughes, 2013). That is why speaking is emphasized among the L2 language learners (Kim & Craig, 2012). In teaching L2 speaking, EFL teachers and course books rely on various approaches, ranging from traditional to modern ones (Eslami, Mirzaei & Dini, 2015; Kim, 2014). Likewise, some other studies (Baker, 2015; Kim & Craig, 2012; Kozulin, 2002) have rarely looked beyond reading and writing skills. Moreover, EFL learners are typically perceived as reticent in class (Sadeghi & Maleki, 2015). A lot of researches (Borich, 2016; Darling-Hammond, 2016; Muijs & Reynolds, 2017; Nilson, 2016; Rivers, 2018) have been done by educational stakeholders to help students gain the required skills. However, teachers cannot rely on some methodologies completely, or more specifically talking, as Kumaravadivelu (2003) acknowledges, “there is no best method there ready and waiting to be discovered” (p. 12). He goes further to believe that it is futile to look for one best method. Accordingly, in the pursuit of the research, the researcher is determined to keep as far away from the old established and prescribed methodologies as he resorts to more interactionist theories such as the one conducted by Long (1985).

A plethora of L2 research (Ghanizadeh, Razavi, & Hosseini, 2018; Sundqvist & Sylvén, 2016; Yang, Chuang, Li, & Tseng, 2013) support the importance of listening-speaking and how comprehensible input facilitates L2 development in the classroom context. Yang, Chuang, Li, and Tseng (2013) argue that developing proficiency in listening is the key to achieve proficiency in speaking.

Two of the highlighted methods in ELT which claimed the development of L2 speaking and meaning negotiation under their guidelines were Audio-lingual Method (ALM) and Task-Based Language Teaching (TBLT). In the present study, this method was operationally defined as the teacher-centered method in which all four skills were sequentially focused on in the classroom and fluency as well as accuracy of the learners' production was of paramount significance. TBLT was also operationalized through the method employed to teach EFL speaking and writing with tasks at its center, the way the theoretical perspectives of this approach have been presented in the literature as well as the way Ellis (2003, 2009) proposed it. Irrespective of the success of both of the aforementioned methods in the Iranian EFL context in the past, the lack of a well-sequenced, centralized, and strong educational method in teaching L2 in general and in the Iranian context, in particular has created a lot of problems for the L2 teaching (Akbari, 2015; Nair, Krishnasamy, & De Mello, 2017).

To fill this educational gap, Halsey (2011) presented her naturalistic-oriented educational proposal, namely the Energizing, Navigating, Generating, Applying, Gauging, and Extending (ENGAGE) Model, in her book titled Brilliance by Design which paved the way for the emergence of educational program changes in America, especially in California where Halsey and Halsey (2017) and Halsey, Halsey, and Gaudette (2018) used the model to develop an educational program stressing the environmental issues in California. Though old traditional and modern methods of language teaching have found their ways to the Iranian educational system (Safari & Rashidi, 2015), to the knowledge of the present researchers, ENGAGE model has not been practiced as a framework in the English Language Teaching (ELT) domain in the Iranian context, yet. Considering the ever-growing demand of Iranian EFL learners for fluent and effective speaking, this study was an attempt to investigate the comparative effects of ALM, TBLT, and ENGAGE model in improving the speaking skill of Iranian EFL Learners.

2. Literature Review

The present section deals with discussing the notions of TBLT, ALM, and ENGAGE model in terms of their teaching L2 speaking methodology.

TBLT

Task-Based Language Teaching was in fact initiated by Prahu (1987) beginning in 1979 (Ellis, 2009). TBLT is an extension of the principles of Communicative Language Teaching (CLT) in which the syllabus is specified in terms of functions and notions. As Ellis (2005) discusses through pre-task planning and within-task planning the advocates of CLT and TBLT focus on real language use in various language skills, especially speaking. Natural learning within the classroom context is one the gifts of TBLT to the learners. There may be cultural barriers to the uptake of TBLT in some parts of the world where people are highly self-culture oriented. Another problem within the scope of TBLT backs to the misunderstanding of the concept of focus on form: Some individuals might think it only pertains to grammar, while it is largely relying on vocabulary as well as pronunciation.

ALM

Combining behavioristic psychology principles and American structural linguistics accompanied with Contrastive Analysis (CA) developments, during and after the World War II, paved the way for the emergence of Oral Approach, the Aural-Oral Approach, and the Structural Approach which later on resulted in the development of ALM (Howatt& Widdowson, 2004). This method emphasized grammatical accuracy in speaking (Richards, 2008). Though ALM is considered the first scientific approach in the ELT (Juffs, 2020), it had its own short comings such as the weak learning theory (Chomsky, 1966) and high amount of meaningless repetitions the ALM used in its instructions (Rivers, 1964).

The ENGAGE Model

Halsey (2011) proposed the concept of Brilliance by Design which was manifested in the ENGAGE model of education. Likewise, Halsey (2016) proposed the idea that individuals’ brain can be energized and their mindfulness, which emphasizes paying deliberate attention to the present moment through observation of thoughts and emotions, without judging can be increased. She asserted that traditional approaches to teaching cannot engage the learner’s mind. To engage the mind Halsey and Halsey (2017) recommend active learning strategies. One such strategy is the ENGAGE Model by Halsey (2011) which “takes a six-step approach to teaching content by using active learning techniques combined with utilizing meaningful interpretation” (Halsey & Halsey, 2017, p. 8). Kim et al. (2017) used the ENGAGE Model in the domain of nursing practices. In this regard, they found “improvement in EBP beliefs had direct effects on improvements in job satisfaction of the participants” (p. 90). Likewise, Glance, Rhinehart, and Brown (2018) have developed a model namely, learn, expand, and engage (LEE) which has been inspired by Halsey’s (2011) ENGAGE Model. They indicate that “the LEE model provides a framework for higher education instruction that directly responds to a recently identified need for competency-based student learning pedagogy in the helping professions” (p. 104).

      Highlighting the environmental issues connected with Chaparral ecosystems and specific and iconic vegetation of California, Rundel (2018) developed an educational program for the operationalization of ecosystem knowledge of the students for turning the tide on urbanization, land-use change and protection of endangered species (p.1). Underwood, Safford, Molinari, and Keeley (2018) also propose that the ENGAGE Model can be used for curriculum development in the educational settings aiming at paving the ground for more awareness toward the environment, wild life, global issues, and consequently more responsible life-long learning. A plethora of research have been conducted on EFL speaking classrooms at the international level (Albino, 2017; Aljumah, 2011; Guchte et al., 2015; Housen & Kuiken, 2009; Jassem, 1997; Kunnu & Sukwises, 2014; Lee, 2009; Nakatani, 2010). Almost all these studies have concluded that EFL students need to be able to overcome the speaking breakdowns resulting from the lack of speaking activities in EFL classes (Ghaemi & Hassannejad, 2015). A key aspect of dealing with such a difficulty knows communication strategies. Therefore, there should be a crucial concern in our English classes for communication strategy instruction with a systematic method. The ENGAGE Model (Halsey, 2011), which takes a relatively new approach to teaching content, could be employed in teaching speaking.

Speaking Components

       Four components of speaking ability that received upsurge attention in the second language research are speaking grammatical accuracy, lexical resources, pronunciation, and fluency. Since 1990s, these four components of speaking have been used predominantly and prominently as dependent variables to assess variation with respect to independent variables such as acquisitional levels or task features” (Sadeghi Beniss & Edalati Bazzaz, 2014, p. 54). In the following section these components are explained in full details.

       Accuracy

      Accuracy is defined as “the capacity of the learner to handle whatever level of inter-language complexity he or she has currently attained” (Skehan, 1998).  Bamanger and Khalid Gashan (2014) also argue that accuracy refers to "how well the target language is produced in relation to the rule system of the target language” (p.4). Further, Skehan postulates that as the language learner tries to produce more accurate speech, he or she does his/her best to take control over the linguistic components that he or she has already acquired. Moreover, Ellis (2019) defines accuracy as “the ability to avoid errors in performance, possibly reflecting higher levels of control in the language and/or a conservative orientation” (p. 545). Housen and Kuikken (2009) consider accuracy as “error-free” speech. Richards and Schmidt (2002) defines accuracy as “the ability to produce grammatically correct sentences but may not include the ability to speak or write fluently” (p.204). Brown (2001) also mentions that accuracy means being “clear, articulate, grammatically and phonologically correct” (p.268). As a component of speaking tests, “grammatical range and accuracy is considered a criterion which focuses on the range and accuracy of the grammar with which the test taker speaks” (Brown, 2006, p. 12). It is mostly about demonstrating confident, accurate sentence construction, including the use of variety of complex structures (with a mix of dependent and independent clauses), all correctly punctuated. It means that not only you need to be accurate in your grammatical use, but you need to use a variety of structures (range).  As it is evident, different scholars suggest different accuracy measures for instance, Skehan and Foster (1997) state accuracy can be measured through percentage of error-free speech, target-like use of vocabulary (Skehan & Foster, 1997); the number of errors per 100 words (Kuiken & Vedder, 2007); and error-free AS-units (Lambert & Engler, 2007).

Fluency

      In scoring speaking tests, “fluency refers to ability to speak smoothly (not quickly) without noticeable effort or loss of coherence. Poor fluency is normally associated with frequent self-correction, hesitation, pausing or repetition” (De Jong, Groenhout, Schoonen & Hulstijn, 2015, p. 223). Coherence, in this regard is embedded in the fluency factor and refers to the unity of the oral text presented (De Jong et al., 2015). In the broad definition, fluency can be seen as overall (speaking) proficiency, whereas fluency in the narrow definition pertains to smoothness and ease of oral linguistic delivery. In this paper, we will use the term fluency in its narrow sense. Ellis and Barkhuizen (2005) state fluency refers to “the production of language in real time without undue pausing or hesitation” (p. 139). Skehan (1998) declares that fluency concerns the learner’s capacity to mobilize an inter-language system to communicate meaning in real time. Furthermore, Richards and Schmidt (2002) state fluency refer to “the features which speech the quality of being natural and normal, including native-like use of pausing, rhythm, intonation, stress, rate of speaking, and use of interjection and interruptions” (p.204). They further assert that in second and foreign language teaching, fluency demonstrate the level of proficiency in L2 communication, which involves:

1. The ability to produce written and/ or spoken language with ease
2. The ability to speak with a good but not necessarily perfect command of introduction, vocabulary, and grammar
3. The ability to communicate ideas effectively
4. The ability to produce continuous speech without causing comprehension difficulties or a breakdown of communication. (Richards & Schmidt, 2002, p.204)

          In addition, Brown (2001, p.268) defines fluency as “basically one’s ability to be understood by both native and non-native listeners”. He also adds that “a higher level would be bilingual, which indicates one is native in two languages, either having learned them simultaneously or one after the other; fluency connotes being flowing and natural” (p.268).  Lyon (2002) states speech Fluency means “the smoothness or flow with which sounds, syllables, words and phrases are joined together when speaking quickly. In this regard ‘fluency disorders’ is used as a collective term for cluttering and stuttering” (p.25).

Lexical Resource

       This scoring category measures L2 vocabulary in the spoken form and the participants will be assessed on their ability to use words appropriately (in the right context) and accurately (with the correct meaning). In some tests like IELTS, test-takers with large vocabularies are best rewarded (Seedhouse & Nakatsuhara, 2018). This criteria focuses on the range of
vocabulary a candidate uses. Generally the wider the range of vocabulary or expression used correctly and appropriately, the better a candidate will score. This means using a more academic style, collocations and understanding connotation, paraphrasing, using vocabulary flexibly, synonyms and word choice (Lexical Resource). These tips will help the candidates to promote their understanding concerning the lexical resource criterion of the speaking tests.

Pronunciation

     As a component of speaking tests, this criterion “focuses on the accuracy and variety of pronunciation features. This includes word stress, which is the stronger pronunciation of a syllable over the others in the word. Stressing the wrong syllable in a word is a frequent error” (Seedhouse & Nakatsuhara, 2018, p. 64). So, what makes pronunciation excellent is intelligibility. If the candidates’ pronunciation interferes with the examiner’s ability to understand them, then they will receive a less-than-perfect score.

      Due to the fact that today in Iranian education system especially in English language institutes, ALM and TBLT are totally common, in this research, these two methods were compared with ENGAGE model. Considering the problems stated above and the purpose of the present study the following research questions were formulated.

• Do ALM, TBLT, and ENGAGE model have statistically significant different effects on the fluency and coherence of Iranian EFL learners in speaking?
• Do ALM, TBLT, and ENGAGE model have statistically significant different effects on the lexical resource of Iranian EFL learners in speaking?
• Do ALM, TBLT, and ENGAGE model have the same statistically significant effects on the grammatical range and accuracy of Iranian EFL learners in speaking?
• Do ALM, TBLT, and ENGAGE model have statistically significant different effects on the pronunciation of Iranian EFL learners in speaking?

3. Research Method

        Participants

The participants of the study were 100 female intermediate level learners (n=100) with the age range of 18 to 25 in one of the language institutes in Damghan (Mofid Language School). These participants were randomly selected out of 150 intermediate students (N=150) attending English conversation classes. Also, their performance in a standard QPT was taken into consideration for the purpose of homogeneity. The selected students were divided into three groups; (ENGAGE model, n=35), (ALM group, n=32), (TBLT group, n=33).

Instrumentation

1. Quick Placement Tests

In order to check the homogeneity of the participants, a standard Oxford Quick Placement Tests (QPT) was used. The test was reported to have had a high reliability (α =.91) based on Cronbach's alpha (Berthold, 2011, p.674. In the present study, the results indicated that QPT had a reliability index of KR-21=.72.

2. IELTS Pretest of speaking

The second instrument used in this study was a standard pretest of speaking selected out of the standard IELTS series. This was done to tap the learners’ L2 speaking knowledge more appropriately based on a standard measure. An "inter-rater reliability index reported for the speaking test of IELTS was (r (3000) = .87, P < .05)" (O'Sullivan, 2018, p.1). This index as O'Sullivan mentions belongs to March, 2018 from over 140 countries worldwide.

3. IELTS Posttest of speaking

The speaking posttest was a new speaking IELTS test selected out of the standard IELTS series. The inter-rater reliability index reported for this speaking test of IELTS was (r (2000) = .82, P < .05)" Fernandez, 2018, p. 8). To score the participants' performance both in the pretest and posttest phases the IELTS Speaking band descriptors (public version), which covered the speaking sub skills was used.

Procedure

       First, the standard QPT was administered to 150 intermediate students. Based on the scale presented for scoring QPT, 100 learners whose scores fell between 24 and 47 were selected as the main participants of the study. The selected participants were randomly assigned to three groups (the ENGAGE group as the experimental group and TBLT as well as ALM as the other groups) with 33 to 34 students in each. Hence, one group received instruction through ALM, another one through TBLT, and the third group received ENGAGE model instruction.

       In the second phase, the participants took part in a pretest of speaking to assure their homogeneity in terms of speaking components. Following the processes of subject selection and getting ensured of the participants’ speaking homogeneity not only in terms of overall speaking skill, but also in terms of speaking components, the researchers launched the intervention phase which lasted 10 sessions. The whole semester included 8 weeks and the learners attended the class three days a week each session lasting for 90 minutes in all groups. It is worth mentioning that the classes of three groups (ALM, ENGAGE and TBLT) received the same hours of instruction and practiced with the same teacher in all groups.

                In the Audio-Lingual Method Group (ALMG), the researcher provided the instruction advocating the principles of ALM which emphasized the use of grammatical sentence patterns. The procedure was as follows: (1) the language teacher gave a brief summary of the content of the dialogue, (2) the language learners listened attentively while the teacher read or recited the dialogue at normal speed several times, and (3) the language learners recited the dialogue line by line or together depending on their length. If the teacher detected an error, it were corrected and the student was asked to repeat the sentence, (4) repetition was continued with groups decreasing in size, (5) pairs of individual acted out of the dialogue. By this time they had been supposed to memorize a text.

      The TBLT group in the present study was exposed to real-world language. An example goes as follows:  The teacher used pictures to elicit learners’ speech and such pictures might also focused on learners’ real-world language and real life issues. Therefore, for the present study, the use of pictures to elicit learners’ speeches was one of the appropriate methods. The students looked at the pictures and spoke about them. They were asked to connect them to their real life situations or bring their own family pictures to the classroom and talk about them. They used photos published in the recent newspaper about a specific novel event, like an accident or a festival, and talked about that. The teacher did not interrupt them while they were speaking. Nor did she fine-tuned their production. This resulted in a less stressful situation for the learners.

      In the experimental group (the ENGAGE Model group), the teacher used the principles of the ENGAGE model (Halsey, 2011). This model employs active learning strategies through naturalist education programs to engage the mind (Halsey, 2016). Hence, the six step general perspectives proposed by Halsey (2011, 2016) were taken into consideration in a language classroom at the intermediate level.

Step1: Energizing students at the beginning of any classroom session through making them involved in the warm ups, ice-breaking discussions, talking about daily life issues, and motivating them through using gestures and postures.

Step 2: Asking the students to navigate what they have gained in the energizing session and develop the new content. . This way the content of what was being taught was developed by the learners and the teacher monitored them to talk about their own interests and concerns.

Step 3: Helping students generate personal meaning and connect what they have gained to their own life and what they feel given the new concepts they have learned and the topic(s) discuss in the classroom. This was done through asking the students to present oral reports to the classroom about the current events, their life and their feelings about recent events in the immediate social context and the like.

 Step 4: Helping students apply their learning to the real world. This was done via asking the students study about the topic selected in the classroom, use the internet, get involved in the social media, collect information about a specific issue, and then present their own perspectives in the classroom. In the next step, students focused on what they could do to bring about a positive change in the social context and their own life.

Step 5: Making leaners gauge and celebrate their progress. This was possible though employing self-assessment (SA) in the classroom context.

Step 6: Helping students extend their learning to action. This became possible through asking the students to use what they had learned in speaking about different issues, lecturing about various topics, taking part in debates and discussions in English and if possible used what they had learned in the social media to find international friends, watch films, and solve the daily life issues and enjoy living through the English language world.

          After the intervention, the three experimental groups received a speaking posttest. This was a new version of IELTS speaking test with a hope to measure the participants’ probable development in speaking sub-skills. An inter-rater scoring system was used to score the learners’ performance in the speaking test and then the inter-rater reliability of the scores in different components was taken into consideration.

4. Findings

          The four research questions of the study addressed L2 speaking components of the learners including fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation of Iranian EFL learners in speaking. A multivariate ANOVA (MANOVA) was run to compare the three groups’ means on the posttests of fluency/coherence, grammatical range and accuracy, lexical resources and pronunciation in order to probe the research questions. Besides assumption of normality which was confirmed in the present study, MANOVA has two specific assumptions; i.e. homogeneity of covariance matrices and homogeneity of variances. The assumption of homogeneity of covariance matrices, which is tested through the Box’s statistics, requires that the correlations between any two tests be roughly the same across the two groups. Table 1 displays the results of the Box’s. The significant results (Box’ M = 38.34, p = .015 > .001) indicated that the assumption of homogeneity of covariance matrices was retained. As noted by Field (2018, p. 972), “Box’s test looks at the assumption of equal covariance matrices. This test can be ignored when sample sizes are equal because when they are, some MANOVA test statistics are robust to violations of this assumption. If group sizes differ this test should be inspected. If the value of Sig. is less than 0.001 then the results of the analysis should not be trusted”. That was why Box’s test was reported at .001 level.

MANOVA also requires that the groups’ variances be roughly the same. Based on the results displayed in Table 2, it can be claimed that the assumption of homogeneity of variances was met on posttests of speaking sub-skills of;

• Fluency and coherence (F (1, 97) = 1.97, p = .144),
• Lexical resources (F (1, 97) = .532, p = .589),
• Grammatical range and accuracy (F (1, 97) = .482, p = .619), and
• Pronunciation (F (1, 97) = .122, p = .300).

Table 3 displays the main results of the MANOVA. Based on these results (F (8, 190) = 18.59, p = .000, Partial η² = .439 representing a large effect size) it can be concluded that there were significant differences between the three groups’ overall means on the four sub-skills of posttests of speaking.

Based on the results displayed in Table 4 and Table 5 it can be claimed that:

• There were significant differences between the TBLT (M = 6.97), ENGAGE (M = 11.40) and ALM (M = 5.65) groups’ means on the posttest of fluency and coherence (F (1, 97) = 203.28, p = .000, Partial η² = .807 representing a large effect size). Thus, the second null-hypothesis was rejected.
• There were significant differences between the TBLT (M = 8.27), ENGAGE (M = 12.25) and ALM (M = 6.90) groups’ means on the posttest of grammatical range and accuracy (F (2, 97) = 119.85, p = .000, Partial η² = .712 representing a large effect size). Thus, the third null-hypothesis was rejected.

• There were significant differences between the TBLT (M = 8.21), ENGAGE (M = 11.14) and ALM (M = 6.90) groups’ means on the posttest of lexical resources (F (2, 97) = 66.59, p = .000, Partial η² = .580 representing a large effect size). Thus, the fourth null-hypothesis was rejected.
• There were significant differences between the TBLT (M = 6.81), ENGAGE (M = 10.77) and ALM (M = 5.59) groups’ means on the posttest of pronunciation (F (2, 97) = 248.81, p = .000, Partial η² = .837 representing a large effect size). Thus, the fifth null-hypothesis was rejected.

            Table 6 displays the results of the post-hoc Scheffe’s tests.

Based on these results and the ones displayed in Table 6 it can be concluded that:

• The ENGAGE group (M = 11.40) significantly outperformed the TBLT group (M = 6.97) on posttest of fluency (Mean Difference = 4.43, p = .000).
• The ENGAGE group (M = 11.40) significantly outperformed the ALM group (M = 5.65) on posttest of fluency (Mean Difference = 5.74, p = .000).
• The TBLT group (M = 6.97) significantly outperformed the ALM group (M = 5.65) on posttest of fluency (Mean Difference = 1.31, p = .000).
• The ENGAGE group (M = 12.25) significantly outperformed the TBLT group (M = 8.27) on posttest of lexical resources (Mean Difference = 3.98, p = .000).
• The ENGAGE group (M = 12.25) significantly outperformed the ALM group (M = 6.90) on posttest of lexical resources (Mean Difference = 5.35, p = .000).
• The TBLT group (M = 8.27) significantly outperformed the ALM group (M = 6.90) on posttest of lexical resources (Mean Difference = 1.37, p = .002).
• The ENGAGE group (M = 11.14) significantly outperformed the TBLT group (M = 8.21) on posttest of grammatical range and accuracy (Mean Difference = 2.93, p = .000).
• The ENGAGE group (M = 11.14) significantly outperformed the ALM group (M = 6.90) on posttest of grammatical range and accuracy (Mean Difference = 4.24, p = .000).
• The TBLT group (M = 8.21) significantly outperformed the ALM group (M = 6.90) on posttest of grammatical range and accuracy (Mean Difference = 1.31, p = .004).
• The ENGAGE group (M = 10.77) significantly outperformed the TBLT group (M = 6.81) on posttest of pronunciation (Mean Difference = 3.95, p = .000).
• The ENGAGE group (M = 10.77) significantly outperformed the ALM group (M = 5.59) on posttest of pronunciation (Mean Difference = 5.18, p = .000).
• The TBLT group (M = 6.81) significantly outperformed the ALM group (M = 5.59) on posttest of pronunciation (Mean Difference = 1.22, p = .000).

5. Discussion

          The first finding of the study was that compared to the TBLT and ALM methods, the ENGAGE model had a more statistically significant effect on the English language speaking development of Iranian EFL learners. As this is the first time the ENGAGE model has been used in the domain of L2 classroom, no previous studies exist in this regard. However, the implications of ENGAGE model in other disciplines and the window it has opened to the new scientific horizons can be discussed here and now. Then, the six steps of ENGAGE will be taken into consideration and the findings of the study will be discussed with regard to the notions and concepts ensued from those steps. 

The first point worth mentioning is that priority of ENGAGE based L2 speaking model over the TBLT which is one of the most successful methods in the ELT domain (Ellis, Skehan, Li, Shintani, & Lambert, 2019), is stunning and attractive. This indicates that irrespective of its novelty and lack of a theoretical linguistic background in this teaching approach, ENGAGE moel has been successful.  

Having been inspired by Halsey’s. (2011) ENGAGE model, Kilbourne (2011) developed his own model for improving safety training which relied on the three notions of connect, inspire, and ENGAGE. In his view, trainers need to spend more time thinking about how they’re going to teach than what they are going to teach. In fact, implicitly he refers to the significance of metacognitive strategies (Novak, 1990; Oxford, 1989). In this regard, the present study findings could find support in the learning psychology operationalized in the preplanning of activities before training the learners. Trainers tend to spend 70 percent of their time focused on what they’re going to teach and only 30 percent on how they’re going to teach it. But Halsey says that it should be the other way around. Halsey et al. (2018) highlight the educational aspects of ENGAGE model asserting that the six steps of the model can pave the way for the success of the learners. They indicate that the ENGAGE Model which has been designed based on discoveries in neuroscience can stimulate active learning and increase retention (Kilbourne, 2011). Since neuroscience and cognition are interwoven and cognitive neuroscience as the scientific field concerning with the study of the biological processes and aspects that underlie cognition pays special attention to the neural connections in the brain which are involved in mental processing, it can be safely concluded that learning in the ENGAGE model enjoys a cognitive orientation. Therefore, it can be assumed that L2 speaking development of the participants has been affected by their cognition and metacognitive strategies operationalized in the ENGAGE model.

In terms of educational significance of ENGAGE Model Rundel (2018) signifies that ecosystem issues should be operationalized in the educational systems and any area and its global significance should be first recognized by the students. It is likely that EFL learners not only improve their L2 abilities through paying attention to such concepts as those of the environment and social life, but also they learn how to connect what they read and learn to the immediate social or environment context.  The success of ENGAGE model in the present study can take support from Kim et al.’s (2017) study on the impact of using ENGAGE model in the domain of nursing practices which found benefits of a regional evidence‐based practice (RBP) fellowship program. In this regard, they found that “improvement in EBP beliefs had direct effects on improvements in job satisfaction of the participants” (p. 90). It is assumed that the L2 learners taking part in the present study also benefited from navigating content and connecting their learning to the real-life situations.

Likewise, ENGAGE model has inspired the development of a model for teaching clinical skills in the helping professions namely, learn, expand, and engage (LEE) (Glance, Rhinehart, & Brown, 2018). LEE has been drawn upon foundations in constructivist philosophies, learner-centered and flipped-classroom pedagogies. On the other hand this teaching model relies on Bloom’s taxonomy and its highest levels. It can be argued that the principles presented in LEE supported by ENGAGE model might be found useful, should they be used in the EFL classroom. 

It is crystal clear that L2 speaking takes the responsibility of facilitating communication between interlocutors. L2 development has also been notified as a life learning concept (Leki, 2017). From this perspective the present findings are in line with another study conducted by Underwood et al., (2018) which has been inspired by the ENGAGE model. They proposed that ENGAGE model can be used for curriculum development in the educational settings aiming at paving the ground for more awareness toward environment, wild life, global issues, and consequently more responsible learning and getting prepared for life learning. ENGAGE Model which proved effective in L2 speaking development can be discussed in terms of its steps and their operationalization in the EFL domain. 

            Energizing learners, as the first step, involves getting learners focused on and excited about training in advance (e.g., having a podcast on the topic, distributing related materials and study guide) (Kilbourne, 2011). At the start of the session, energizing also includes thanking learners for participating and getting them involved right away by asking a powerful opening question, conducting an interactive activity, or announcing key training goals (Halsey et al., 2018). In the L2 speaking classroom warm ups, ice-breaking discussions, talking about daily life issues, and motivating students through using gestures and postures were taken into consideration. Such notion can find support in eth EFL classroom literature (Sert, 2015; Scrivener, 2012). Also, this step deals with motivation; both internal and external, which have been researched concerning their effectiveness in L2 development (Csizér, 2017; Ushioda & Dörnyei, 2017).

Navigating content, the second step in the ENGAGE model focuses on using a variety of methods (e.g., visual, auditory, kinesthetic) to engage different parts of the brain, the trainer alternates between teaching content and review content through role-plays, games, or team activities (Halsey, 2018). In the L2 speaking class, asking the students to navigate what they have gained in the energizing session and develop the new content was of paramount significance, Likewise, the teacher and learners negotiated on decisions to be made about assignments and activities. This indicates the application of process-based syllabus (Breen, 1987) and negotiated syllabus (Clarke, 1991) in the EFL pedagogy.

Generating meaning, as the third step, urges the learners to clarify the value of the new information they learned and how it will help them; for example, learn more successfully and diagnosing the problem while learning (Halsey, 2011). In the L2 speaking class, this step was operationalized through asking the students to present oral reports to the classroom about the current events, their life and their feelings about recent events in the immediate social context and the like. This is partially in line with TBLT principles proposed by Ellis (2003), especially the real language tasks.

           Applying to the real world, as the fourth step, signifies that, learners need opportunities during the teaching/ learning process to demonstrate their mastery of the new skills (e.g., learning pronunciation, intonation, lexical resources, or real-world practice). In the L2 speaking class, this notion was implemented through asking the students to study about the topic selected in the classroom, use the internet, get involved in the social media, collect information about a specific issue, and then present their own perspectives in the classroom. Such tasks have been employed in action research (Smith & Rebolledo, 2018). Gauging and celebrating, as the fifth step of ENGAGE model concentrates on learners’ assessing their own learning and development and how much they have learned-through a quiz, crossword puzzle, or presentation to others-and celebrate their accomplishment. This concept was operationalized by employing teaching self-assessment (SA) principles and how to develop SA speaking checklists in the classroom context. 

Extending learning to action, as the sixth step of ENGAGE model pertained to follow-up activities (e.g., e-mail reminders or buddy systems) to help ensure that learners act on their intentions to use their new knowledge or skills (Halsey, 2011). This step was applied to the L2 speaking class through asking the students to use what they have learned in speaking about different issues. They were encouraged to talk about various topics, take part in debates and discussions in English and if possible, use what they have learned in the social media to find international friends, watch films, and solve the daily life issues and enjoy living through the English language world. This is in line with competency-based learning in the ELT domain (Nodine, 2016; Waddington, 2017).

6. Conclusion

          The quantitative data analysis provided the searcher with two sets findings: a) there were significant differences between the TBLT, ENGAGE, and ALM groups’ means on the posttest of speaking. The ENGAGE group significantly outperformed both the TBLT and ALM groups on posttest of speaking. Likewise, the TBLT group significantly outperformed the ALM group on posttest of speaking. b. There were significant differences between the TBLT, ENGAGE, and ALM groups’ means on the posttests of fluency and coherence, lexical resources, grammatical range and accuracy, and pronunciation. The ENGAGE group significantly outperformed the TBLT and ALM groups on the posttests of fluency and coherence, lexical resources, grammatical range and accuracy, and pronunciation, while the TBLT group significantly outperformed the ALM group on the same posttests. Another point of divergence could be focusing on the grammatical accuracy in assessing L2 speaking. The ALM group mainly stressed on the grammatical errors, while for the ENGAGE and TBLT groups, pronunciation and lexical resource were more important and this was followed by fluency. To sum up, the results of the present study showed that learners experiencing ENGAGE model teaching techniques resulted in better speaking commands compared to their counterparts receiving TBLT or ALM.  This way the method can leads to higher strategic L2 speaking development in an EFL context. Hence, it postulates that ENGAGE based teaching techniques provide a better learning context for EFL learners’ L2 speaking compared to those of the TBLT and ALM methods.