Classroom Interaction as a Way of Developing Students’ Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation

Teaching speaking skills through classroom interaction is usually a neglected program at the tertiary EFL education in Bangladesh. Th is study primarily aims at discovering the current scenario of the practices and problems of classroom interaction at a Bangladeshi tertiary EFL classroom. Th is study further proposes some strategies for developing speaking skills through classroom interaction. Th e data were collected from 110 student respondents and 11 teacher respondents of 4 universities based on the stratifi ed random sampling. For collecting and analyzing data, a mixed-methods approach (QUAN-QUAL) was applied. Th e data collection methods were a questionnaire survey on the tertiary EFL students and a semi-structured interview of the tertiary EFL teachers. Th e results show that though students and teachers were aware of classroom interaction, very little communication actually took place in the classroom because of teachers’ monopolizing the talk time and learners’ getting little to no talk time at all. Th e study also exposes that most of the students did not interact spontaneously and teachers used a great deal of Bangla in the EFL classroom. Th ese hampered students’ speaking skills development. Finally, this paper presents some pedagogical implications and off ers some recommendations for the considerations of the tertiary EFL teachers, learners, and policymakers.


Introduction
In the current world, English is the most prominent lingua franca which is massively used at international business meetings and academic conferences (Formkin, Rodman, & Hyams, 2011). Robson (2011) states "English is spoken at a useful level by some 1.75 billion people worldwide -that's one in every four" (p. 2). Currently, it has become the most dominant language in the fi elds of trade and commerce, science and technology, international relations, and education (Harmer, 2011) and over 1 billion people across the globe are learning English (Luke, 2021). To cope with this reality, ELT practices in Bangladesh had a shift from the popular In another study conducted on the humanities students of Dhaka University, Chowdhury (2010) finds that more than 50% of the students "rarely-never" (p. 69) spoke in the classroom. Even though from her study we find a picture of classroom interaction and the state of our tertiary students' speaking ability, the data has become somewhat old which was sampled from only one university. Thus, the researchers of the current study got interested in recognizing the current general state of the tertiary students' speaking proficiency and the practices of Bangladeshi classroom interaction. Alam and Khan (2014) and Chowdhury and Kabir (2014) view that Bangladeshi tertiary students are usually weak in communicative skills. In this regard, they find some gaps between the theory and the practice in teaching English communication skills, particularly speaking but they have not mentioned explicitly what those gaps 222 CROSSINGS: VOL. 12 | 2021 | ISSN 2071-1107 Classroom Interaction as a Way of Developing Students' Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation are and how much the students are weak in interactive speaking. Therefore, the present researchers became curious to unearth the gaps and reasons for students' weaknesses in this skill.
Taking all the realities into consideration, the researchers of the current study decided to conduct a study and the primary objective of it is to observe the real classroom interaction practices, particularly the role of teachers in the tertiary EFL classroom of Bangladesh. Another objective of the study is to show how far classroom interaction strategies develop learners' speaking ability. Finally, it aims to explore the gap between the standard practice of classroom interaction and the real classroom interaction scenario here at the tertiary level EFL classroom and the obstacles EFL teachers and learners face in carrying out interactive activities in the language classroom.
In order to meet the objectives discussed above, three research questions were formulated to carry out the present investigation. The research questions are given below:

Research question 1:
What is the common classroom interaction scenario in the tertiary EFL classes, and how far does classroom interaction enhance tertiary EFL students' speaking skills?

Research question 2:
What roles are played and strategies followed by the tertiary EFL teachers of Bangladesh to make the EFL classes interactive and develop students' speaking skills?

Research question 3:
What hindrances do tertiary EFL teachers and students encounter while carrying out interactive activities in a speaking class?
For obtaining the answers to the research questions, a brief survey was conducted in four universities of Dhaka that helped the researchers elicit the views of the concerned students and teachers. The survey results have been presented and discussed in this paper systematically. Finally, based on the findings and analyses, some pedagogical implications have been highlighted for the concerned stakeholders: students, teachers, and policymakers.

Background of the study
The evolution of ELT in Bangladesh: Past and present Though English in Bangladesh is generally considered a foreign language, McArthur (1996) terms Bangladesh as an ESL country. In this connection, Sarwar (2013) states, "English has the status of an unofficial second language in Bangladesh and is compulsory from the primary level in all state-run schools where the medium of instruction is Bengali" (p. 145). In the Pakistan period, English enjoyed the status of the second language and it was taught as a functional means of communication at secondary schools in Pakistan (Hasan 2004) but immediately after the liberation in 1971, the status of English was changed and Bangla was given higher prestige in all spheres of the society, predominantly in education (Rahman et al, 2019). The outcome was not beneficial to the EFL learners here as it led to a fall in the proficiencies in English skills among the learner communities (Rahman et al, 2019). As part of the policy, English as a compulsory subject was withdrawn from the tertiary level. Moreover, because of the Bengali Language Implementation Act 1987, English lost the status of the official second language which yielded a drastic result as "the standard of English fell to the abysmal depth in public schools and universities" (Hassan, 2004, p.11). Practically we fell behind although, ideally, we wanted to proceed.
Having realized the consequences of students' low performances in English skills, the government later came forward to make up for the loss. In order to develop students' communicative competence, CLT was first introduced in the secondary and higher secondary levels in the 1990s by the National Curriculum and Textbook Board (Barman, Sultana & Basu, 2007). The Private University Act 1992 prompted the setting up of a good number of universities and the act allowed English to be used as the medium of instruction. According to the University Grants Commission of Bangladesh (UGC), there are 49 public universities and 107 private universities operational in seven divisions of the country (UGC, 2021). For all disciplines, Functional English courses have been made obligatory by the UGC and the latest education policy, National Education Policy-2010 (National Education Policy 2010: p. 32, Clause 09). Many universities especially private universities now offer additional one or two English courses of 100 marks for 3 credits in order to give extra support to the development of students' English language skills (Akteruzzaman & Islam, 2017). They have initiated different EFL programs to teach students the basic language skills (BLS) of listening, speaking, reading, and writing in English and prepare them for the global competitions and job opportunities (Sultana, 2014;Choudhury & Kabir, 2014). These initiatives have already helped the learners excel in academic English exams and the passing rates in these subjects have increased notably though unfortunately, these rates do not reflect the original development of their language skills (Chowdhury & Kamal, 2014).

Literature Review
Teaching interactional speaking skills at the EFL classroom: Pedagogical suggestions "Classroom interaction has been a central issue in teaching and learning English in the era of communicative language teaching" (Sundari, 2017, p.146 (Ellis, 1991;Krashen, 1989;Larsen-Freeman & Long, 1991) have added significant contributions to the development of language teaching and learning. English being a major international language of communication, it is important that EFL learners learn to speak and interact in a multiplicity of situations using the language. Especially, speaking in English is considered a skill upon which a person is judged "at face value" (McDonough, Shaw, & Masuhara 2013, p. 157). Therefore, the intricacy of speaking and interactional skills can be described by "…the ability to fill time with talk…the ability to talk in coherent, reasoned and semantically dense sentences… the ability to have appropriate things to say in a wide range of contexts" (Richards, 1990, p. 75).
Regarding the interactional activities in the EFL classroom, Shumin (2002) suggests that primarily adult learners should be engaged in short, interactional exchanges in which they will make short utterances and then participate in small talks in the target language and so on. Pair work and group work are also suggested as important strategies to increase speaking fluency (Brown, 2001;Green, 1989;Nation 1989). Nation (1989), for instance, affirms that "one of the most useful procedures is the movement from individual to pair to group to whole class activity" (p. 26). Regarding teaching interactional skills in the speaking class, Bohlke (2014), Goh (2007) and Nunan (2015) also put emphasis on undertaking interactional skills like regulated turn-takings, maintaining and closing a dialogue, and other conversation management strategies alike.
In order to enhance classroom interaction, Nunan (2015) suggests, while designing the speaking skill syllabus, teachers should consider the global, national, and local contexts and create opportunities for students to talk by using group work or pair work. Tomlinson and Dat (2004) also suggest that teachers should encourage oral participation, nurture a supportive atmosphere in the classroom, provide constructive feedback to the production of students, and encourage peer interaction in the speaking development classes.

Classroom interaction and the development of students' speaking skills: Teachers' role in the EFL classroom
In the EFL classroom, non-native English speakers need to be prepared to speak in different situations and classroom interactions can be a strong way to develop their oral skills. In order to facilitate students' speaking fluency, teachers should ensure their maximum exposure to the target language (Brown, 2007;Scrinvener, 2005) and motivate them to interact with each other (Ur, 1996). Regarding the types of interaction, Rivers (1987) and Tomlinson and Dat (2004) suggest two major kinds: teacher-learner interaction and learner-learner interaction. They also suggest that teachers must make a practical arrangement in the EFL class so that students get enough opportunities to think, reflect and rehearse.
With regards to developing learners' speaking abilities, Harmer (2011) stresses that in the EFL classroom a language teacher should play the roles of a facilitator, a manager and an organizer of different interactive activities like: designing and organizing pair work and group work, giving feedback, playing whatever role that is required, and arranging different spoken discourse activities. One of the major problems EFL learners face in the classroom is the anxiety to interact which has a negative impact on their oral fluency (Horwitz, Cope and Horwitz, 1986). In order to help students overcome this anxiety, Brown (2001) suggests, "Our job as teachers is to provide the kind of warm, embracing climate that encourages students to speak, however halting or broken their attempts may be" (p. 269).

Studies on interactional speaking activities at the tertiary EFL classroom
In a research on the tertiary EFL students in Palestine, Nakhalah (2016) finds "the levels of students of English at Al Quds Open University in speaking skills is medium" (p. 104). He also observes that the lack of a proper interactional environment in the EFL classroom and teachers' negligence in arranging classroom interaction are major reasons for learners' backwardness in improving their speaking skills.
In another research conducted on the Tunisian tertiary teachers, Ounis (2017) reports that Tunisian EFL teachers carry out diverse types of classroom activities like "oral presentations, debates and role plays as the major speaking tasks" (p. 101) and motivates learners to be involved in communicative interactions with one another. About the output of the classroom interaction and oral assessments, she expresses her satisfaction and states "Therefore, the careful choice of oral performance tasks mirrors Stiggin's (2008) ideas that assessment does help teachers make instructional and educational decisions. Through the classroom management of appropriate and relevant tasks, the quality of assessment is guaranteed" (Ounis, 2017, pp. 101-102).
In another similar study conducted on some private universities in Bangladesh, Farooqui (2007) finds tertiary students' spoken proficiency is very poor, although private universities provide students with various interactive activities and competitions. She further observes that most of the students feel shy and do not want to interact in front of the class. Mridha and Muniruzzaman (2020) find a similar scenario in public universities where "a huge number of students never talk in the class" (p. 131). Regarding teachers' initiative to promote classroom interaction, they find that many of the teachers do not care about their responsibility to promote conversational skills in the EFL classroom and "such kind of non-cooperation on the part of teachers highly contributes to the lacking of competency in conversation skill in English" (Mridha & Muniruzzaman, 2020, p. 131). Seedhouse and Jenks (2015) suggest that in the language classroom, interaction should be made an inseparable part of education. In our case in Bangladesh, the situation is different and there seems to be a big gap between policy imperatives and classroom realities; thus it can be concluded that here the tertiary level language skills and curriculum objectives are rather poorly met (Rahman et al., 2019).
The literature on teaching interactional spoken English to tertiary learners in Bangladesh is very scarce (Farooqui, 2007). In addition, regarding the practices of speaking skills like how much time is spent for developing the interactional speaking skills, what roles teachers play in the classroom, what strategies they follow in carrying out the interactive activities at the tertiary EFL classroom and so on, very little data is available. From the available data it is found that there are some practices of speaking skills in the classroom but in many cases the results are not up to the mark. Though we get some information regarding classroom practices at private universities from Farooqui (2007), Mridha and Muniruzzaman (2020), the studies were conducted in very limited research sites and hence the findings cannot be generalized.
Regarding the interactional speaking skills practices in the public universality EFL classroom, Mridha and Muniruzzaman (2020) report a very grim picture. In their study they find that about 40% of the students very rarely interact in the class, 53% students interact occasionally and only 8% students interact in English very often. From the studies of Maniruzzaman (2012), and Alam and Sinha (2009) regarding teaching and learning speaking skills, it is further proved that indeed, speaking is a neglected skill at the tertiary level of Bangladesh. However, there is not much clear and comprehensive data available on how much this skill is neglected in the local instructions and assessment system and what interactive activities are undertaken by the tertiary EFL teachers. Therefore, additional and updated data need to be collected to enable the researchers to examine and address the issue more closely and extensively.

Research design
In terms of approach, it was a mixed-methods research as both the aspects of quantitative and qualitative methods in the stages of data collection and data analyses were employed. In terms of typology, it is a descriptive and explanatory research in nature.

Study areas
The proposed study areas were 4 selected private universities in Dhaka Metropolis. For collecting the necessary data, researchers visited those universities for about 25 days starting from 29 th October to 22 nd November 2019. For visiting the universities and collecting data from the students and teachers, permission from the heads of the departments of English and the concerned authorities of the universities were taken in due manner.

Participants of the study
The target population/subjects for the present study were of the tertiary levelundergraduate students and their teachers. The student respondents were university students who were either studying or had already studied compulsory functional English courses at the entry-level and the teacher respondents were their concerned English language teachers.
In total 110 students from 4 Universities based on stratified random sampling were selected, and they participated in the questionnaire survey voluntarily. Among the student respondents 79 were male and 31, female. Their ages werebetween 20-25 years.
A total of 11 EFL teachers participated in the research. Among them 5 were male and 6 were female. All of them graduated in English from different universities and most of them were highly experienced English Teachers. The total sample size of the study was 121.

Data sources
The cardinal sources of our data were the students and the EFL teachers of the departments of English from the universities we selected. Students' feedback in questionnaire surveys and interview feedback of their EFL teachers were the most important primary data. Curriculum development theory books, books on language skills theories and research methods, government gazettes, relevant MPhil and PhD dissertations, and internet websites were used as the secondary sources of data.

Data collection instruments
The instruments used to collect the data were students' questionnaire for questionnaire survey, interview notes, and note-taking sheets for semi-structured interviews of the teachers.
The student questionnaire consisted of 11 questions with multiple choice answers. For the purpose of maintaining validity and reliability, the questionnaire contained questions that directly matched the major issues of investigation of the study. Thus, through the student questionnaire our quantitative data were collected from the students. A similar instrument was used in the studies of Mridha and Muniruzzaman (2020), Tercan and Dikilitaş (2015), and Ounis (2017).
Later, in order to find out the gaps between students' expectations and their achievements regarding the improvement of their speaking skills, the researchers took a semi-structured one-to-one interview of some EFL teachers of the concerned universities. Thus, our qualitative data were collected from the teachers. To ensure the authenticity of the data, due permission was taken from the authorities and the teachers beforehand while the confidentiality of the respondents' personal 228 CROSSINGS: VOL. 12 | 2021 | ISSN 2071-1107 Classroom Interaction as a Way of Developing Students' Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation information was guaranteed. The interviews followed a semi-structured format because the study started with a fairly clear focus and then it addressed more specific issues. Each question was followed by inquiries planned to attain more thorough replies. Each interview lasted about 35 minutes and they were conducted in the participants' university offices. Most of the questions focused on the strategies that teachers used in teaching interactive speaking skills in the EFL classroom. Some questions were asked about their perceptions regarding students' classroom interaction difficulties and the steps teachers took to minimize them (the interview questions are given in Appendix B). Interview transcripts were later sent to the respondents for verification to increase the credibility of the study. Fade (2003) and Mays and Pope (2000) followed a similar technique in their research. In order to collect similar data, this instrument was also used in similar studies conducted by Farooqui (2007), and Nakhalah (2016).

Data collection procedure
To collect data from the student respondents the researchers visited their EFL classes at the appointed times and one of the faculty members of each university helped the researchers to meet the student respondents. Then the researchers briefed those students about the purpose, objectives, and significance of the current study and clarified all their queries. Finally, the respondents were provided with the questionnaires. After filling out the questionnaires, they returned the papers to the researchers. This way the quantitative data were collected from 110 student respondents of 4 universities.
To collect the data from the teachers, they were explained the reasons and aims of the current research, also the major questions of the thesis. The interviews were recorded and all the important points of their feedback were noted down. All the interviewees were asked the same questions as noted on the interview sheet.

Data analysis and presentation procedure
The current research adopted thematic approaches for data gathering and analysis. All qualitative data were logically interpreted and placed. Mathematical tools and descriptive statistical methods were used for analyzing the quantitative data using MS Excel 19 and were presented in tables, followed by subsequent discussions.

Data Analysis and Findings
Analysis of quantitative data collected from student questionnaire In an attempt to investigate the importance of classroom interaction in improving EFL students' speaking skills and the real scenario of classroom interaction and the relevant challenges tertiary students faced in developing their speaking ability, student respondents were asked 14 multiple choice questions through a questionnaire survey. The data collected from their answers are analyzed below.  The data depicted in Table 1 show that the majority of the respondents (71%) experienced a poor show of student-student interaction inside the EFL classroom while only 29% felt that there was enough opportunity for them to participate in peer interaction. This finding is very parallel to that of Mridha and Muniruzzaman (2020) who find that only "6.66% often and 0.95% respondents always speak English" (p. 125). This goes against the popular suggestion of the scholars regarding the expected amount of student-student interaction that supposedly takes place in our EFL classes. Seedhouse and Jenks (2015) suggest, in the language classroom interaction should be made an inseparable part of education. This picture also goes against the curricular objectives of the tertiary EFL education of Bangladesh. The same is also resonated with the observations of Rahman et al., (2019) who have found evidence that the curriculum objectives of our tertiary level language skills are hardly met in practice (Rahman et al., 2019).  Table 2 shows a poor picture of student-student interaction outside the classroom. The data show that only 26% of the students tried positively to interact outside the classroom while 38% interacted occasionally and a big percentage (36%) rarely interacted outside the classroom. It is observed that the majority of our student respondents were reluctant to interact in English both inside and outside the EFL classroom.

Amount/frequency of student-student interaction outside the classroom
Amount/frequency of student-teacher interaction inside the classroom From the results obtained from Table 3, it is seen that only 40% of the student respondents felt there was enough student-teacher interaction in the EFL classroom, while 39% felt this sort of interaction took place often and occasionally. On the other hand, a significant number of respondents (21%) experienced that student-teacher interaction in the EFL classroom took place very rarely. However, it can be deduced that the rate of teacher-student interaction in the EFL classroom was not satisfactory enough. Although in the Tunisian context it is seen that this support is satisfactory enough in the tertiary EFL classroom (Ounis, 2017), in the Bangladeshi context this picture is different as reported by the study of Mridha and Muniruzzaman (2020). In this regard, they find that many of the teachers do not care about their responsibility to promote speaking skills in the tertiary EFL classroom.  Table 4 shows the majority of the student respondents (65%) felt the most common pattern of interactions practiced in the classroom was "T-Ss" and 45% felt "T-S." That means most of the interactions were teacher centered. On the other hand, on average, 23% of the respondents felt that the patterns "Ss-T" or "S-T" took place too and, an average 12% voted for "S-S" and "Ss-Ss" patterns. This picture clearly shows that student-student interaction was very rare in the tertiary EFL classroom and that the EFL classes were highly teacher-centered. This finding is also supported by the study of Mridha and Muniruzzaman (2020) who report "a huge number of students never talk in the class" (p. 131). They also observe that students "do not feel comfortable speaking English with one another" (p. 131). Regarding student respondents' perception about the most popular interaction techniques and strategies practiced in the speaking skill classroom, Table 5 shows, 55% of the student respondents believed the most effective classroom interaction techniques were pair work and group work. Their perception is also in tandem with that of Harmer (2011) who suggests that a language teacher should be a facilitator and an organizer of different activities like pair work and group work, role plays, different spoken discourses, and interactive activities. This is also supported by the study of Khan (2017) who report that "most students prefer group discussion for problem-solving' (p. 57). Contribution of classroom interaction to the development of students' speaking skills  Table 6 clearly shows that the vast majority of the student respondents (86%) believed that classroom interaction was a useful tool to enhance their oral skills. This finding agrees with the suggestions of Shumin (2002) who observes that graded interactive activities in the EFL classroom will make the learners confident in speaking gradually and will lead them to be successful speakers in the long run. Brown (2001), Green (1989), and Nation (1989), in addition, suggest that pair work and group work are two of the major interactive activities that will help increase the successful chances of speaking fluency in adult learners. Regarding students' satisfaction level of the classroom interaction scenario, the results displayed in Table 7 imply that about one-third of the respondents were happy with the practices of the classroom interaction while two-thirds of them were not satisfied enough. The results show that only 22% respondents felt that their EFL classes were truly interactive while 18% believed that the classroom interaction scenario was just satisfactory enough and the majority of the respondents (60%) said that the classroom interaction scenario was poor. These findings are also supported by the studies of Mridha and Muniruzzaman (2020), Maniruzzaman (2012), and Alam and Sinha (2009). They observe that though Bangladeshi tertiary education has theoretically adopted speaking skill as an important language skill to be practiced in the EFL classroom, it is still a neglected skill in both instructions and assessments. The data of Table 8 show that the speaking proficiency level of the majority of the student respondents (59%) was not up to the mark while the speaking proficiency level of 41% was satisfactory, i.e. good or very good. These findings are synonymous with the findings of Mridha and Muniruzzaman (2020) who report that in the tertiary EFL classes "… a huge number of students never talk in the class" (p. 131). They also observe that tertiary students do not feel comfortable with using English in the classroom too (p. 131).  Regarding teachers' role in encouraging students to develop their interactive speaking skills in the classroom, it was found, as per Table 10 that the majority of the EFL teachers (72%) encouraged and motivated students to interact in English while 28% of the teachers did not act to the expected extent. From the data it is seen that the teachers' motivation level was high but in this case the report of Mridha and Muniruzzaman (2020) gives us something different. They observe that many of the tertiary EFL teachers did not bother about discharging their responsibility as an organizer or a facilitator; rather, they played the role of a controller and a ruler and such a role highly contributed to the students' "lacking of speaking competency in conversation skills in English" (p. 132).  Table 11 shows that most of the student respondents (51% equally) thought their teachers played two main roles: one of a corrector and the other of an organizer. 37% of the respondents gave their opinions about the teachers' role being that of the controller, 35% of the tutor, 33% of the prompter, 31% of the participant, 30% of the resource person, 26% of the observer and 9%, the assessor. Although language teachers play some general roles, to improve students' interactive skills, they should also play the roles of an organizer or a manager in the EFL classroom (Harmer, 2011). In this case, Table 11 shows that only 51% teachers played this role and the rest of them did not, which logically demonstrates that there were much fewer interactive activities in the EFL classroom than expected. On the other hand, it is seen that teachers' greater roles of an organizer and a motivator brought a very good result for the Tunisian tertiary students. In a study conducted on the Tunisian tertiary EFL teachers, Ounis (2017) reports that they successfully carried out diverse types of classroom activities like "oral presentations, debates and role plays as the major speaking tasks" (p. 101) and motivated learners to be involved in communicative interactions with one another. About the output of the classroom interaction and oral assessment with the EFL learners, she reports a very satisfactory result (p. 101). From the data it is seen that regarding classroom interaction Bangladeshi EFL learners, in comparison, had much less of an expected service from their EFL teachers. Results, as presented in Table 12, show that the majority of the teachers (75%) were positively active in correcting students' mistakes/errors or giving feedback in speaking lessons while one-fourth of them were not active in this role. This finding supports what some other researchers, too have discovered. They opine that in the language classroom the teacher usually controls the activities as the knower and gives feedback or correction whenever students make errors (Murray & Christison, 2011;Ur, 2009;Walsh, 2011). From the data of the current research, it is seen that the majority of the teachers were active in helping the students overcome their speaking difficulties and errors. From the findings it can be expected that tertiary students greatly benefited out of their teachers' services but the research results of Mridha and Muniruzzaman (2020) contradict the expectations as they find only 7% of the tertiary students interacted spontaneously in the EFL classroom (Mridha and Muniruzzaman, 2020). When asked about their own perception regarding the reasons for their speaking reticence in the classroom, the majority of the student respondents (86%) said, as presented in Table 13, they lacked enough language resources and confidence to speak and interact in the classroom. This finding is very synonymous with that of Farooqui (2007) who reports that EFL students usually have a small English vocabulary and perform very poorly in speaking. As the reason, she says, "They feel shy and do not want to speak in front of the class" (Farooqui, 2007, p. 102). About one-third of the respondents of the current research also opined that they did not find the lessons interesting or engaging and that their classroom teaching/ learning environment was not supportive enough to enhance their participation in interactive activities. This reason can be attributed to the reasons shown by Farooqui (2007) as she finds that students come to study at the tertiary level in Bangladesh with an educational background where EFL classes see very little interaction and creativity. She further adds that although university teachers try their best to support their students, most of the pupils feel nervous to initiate any interaction because of their weak background in English (Farooqui, 2007

Analysis of Qualitative Data Collected from Teachers' Interviews
In an attempt to investigate the importance of classroom interaction to improve EFL students' speaking skills and the real scenario of classroom communication in the tertiary EFL classes, also to discuss the many challenges our teachers and students face while developing their speaking abilities, a total of 11 English teachers from 4 universities were interviewed. The interview sheet consisted of 11 questions: both "closed" and "open-ended." The closed questions required teachers to answer "Yes" or "No", and through the open-ended questions, teachers' elaborate views and their knowledge about the issues were elicited. During the interview, all the salient points and themes from the teachers' replies to each question were noted down. In addition, all interviews were audio-recorded to support the data analyses where necessary. Each interview lasted approximately half an hour.

General Findings
Prior to the interview, teachers provided information on their personal background through Question no. 1 (see Appendix B). Teachers' information gathered from the bio-data show that there were 5 male teachers and 6 female teachers. Their ages ranged from 22 to 35 years. All of them were Bangladeshi nationals. Nine of them graduated from the University of Dhaka, one from Shahjalal University of Science and Technology, and the other from American International University of Bangladesh.
The data collected from their responses show that 3 teachers had been teaching English for eight years, 1 teacher for five years, 3 for four years, 3 for three years, and the rest 2 for about one year. This denotes that a majority of them were experienced EFL teachers. When asked about their job satisfaction, they gave positive feedback about their profession and responsibilities as an EFL teacher and shared some of their happy memories of their teaching career. Most of them said that although there were some challenges and issues, they liked interacting with the students in the question-answer session. They also affirmed that in their respective EFL classes they shared ideas and experiences with the students in all necessary cases. 3 teachers pointed out that they appreciated the efforts of the students in classroom interaction.

Special Findings
The recurrent themes and significant comments of EFL teachers regarding the teaching of interactive speaking skills in the EFL classroom were identified and incorporated into three main sections: • Roles played and strategies followed by EFL teachers to make classes interactive and develop students' speaking skills • How far classroom interaction enhanced tertiary EFL students' speaking skills, and 238 CROSSINGS: VOL. 12 | 2021 | ISSN 2071-1107 Classroom Interaction as a Way of Developing Students' Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation • The major hindrances teachers encountered in developing students' interactive speaking skills

Roles played and strategies followed by EFL teachers to make classes interactive and develop students' speaking skills
While answering questions regarding the above-mentioned issues most of the teachers said that they were aware of the many benefits of classroom interaction but the actual instances of such communication were much lesser than necessary. They attributed this gap to the shortage of class time. Most of them further reported that the common interaction pattern they followed in the classroom was T-Ss. This implies that the classes were largely teacher-centered. A similar finding is also supported by the work of Mridha and Muniruzzaman (2020) who found that the Bangladeshi tertiary EFL classes were mostly teacher-dominated and an ideal interactive scenario was largely absent in the EFL classroom because of that. The same is also supported by Thorton (2006) who observes that students came to tertiary level in Bangladesh with an educational background where "classroom activity is characterised by the teacher delivering the syllabus which students ingest, leaving little room for genuine enquiry, questioning or criticism" (p. 190). Some of the teachers said they applied S-S, T-S, S-Ss patterns too because they thought that students' speaking skills would develop further if they talked with their classmates more than they did with their teachers.

How far classroom interaction enhanced tertiary EFL students' speaking skills
The data show only 1 out of 11 teachers used mainly English as the medium of classroom instruction in the EFL class. While asked about her motivation behind carrying out the classroom instruction in English, she said that she had studied in an English medium school where English was the primary medium of conversation. She added that she felt comfortable with speaking and teaching in English in her classroom. All the rest of the teachers reported that they used both Bengali and English in the classroom teaching. While asked about the reason for using two languages in the classroom, they gave more or less a common answer. They said if they always interacted with students in English, students would have failed to understand the text properly.
In a close-ended question, teachers were asked if classroom interaction was a meaningful way to develop learners' speaking ability. In reply to the question, every teacher said "yes." In reply to the query regarding the effectiveness of classroom interaction to help students develop their speaking skills, all the teachers opined that their students learned much better if they worked in pairs or groups. They also added that this activity developed students' intellectual understanding, communication skills, managerial qualities, interpersonal skills, etc., and helped them to be more fluent in speaking. Some teachers said students worked better in pair/group tasks as weak students got ideas and support from the stronger ones in shared talks.
In reply to the question regarding what interactive activities they mostly used in the speaking classrooms, the data demonstrate that, most of the teachers organized group work, pair work, role-plays, warm-up sessions and they also used multimedia to facilitate and motivate students to interact in the classroom. The finding agrees with the suggestions of Harmer (2011) who opines that a language teacher should organize different activities like pair work, group work, role-plays, and different spoken discourse or interactive activities in the EFL classroom. This is further supported by the study of Khan (2017) who reports that "most students prefer group discussion for problem-solving" (p. 57).

Major hindrances teachers encountered in developing students' speaking skills
In reply to the question regarding the above-mentioned query, all the teachers reported some obstacles they faced trying to make their classes interactive. The major ones are: students' poor vocabulary stock and other linguistic resources; gaps among students' language proficiency levels; lack of proper academic environment and motivation; poor speaking proficiency level and lack of confidence. The data also agree with the observations of Savaşçı (2014) who finds that feelings such as lack of confidence and fear of making errors are some of the major causes for students' speaking reticence in the classroom (p. 2686). The teachers also reported that short instructional periods might turn out to be another obstacle to the development of students' interactive speaking skills.
Furthermore, it is understood that institutional policies, curriculum, students' exposure to the language, and social and educational backgrounds apparently influence how teachers and students interact with each other. This research finding also supports the suggestions by Stern (1983, as cited in Hall, 2011 about contextual factors in language teaching. Seedhouse and Jenks (2015) also observe that classroom interaction might be affected by the institutional and national policies and learners' linguistic or cultural backgrounds.

Discussions on the Main Research Questions
In the questionnaire survey for students under Q.I -Q.X, respondents were asked to give their own perceptions regarding different practices of classroom interaction they experienced in their EFL classes. Through these questions, they were also asked to comment about the efficacy of the classroom interaction activities in developing their speaking skills. Q.XI -Q.XIII were based on what roles teachers played to make the EFL classes interactive and develop students' speaking skills while Q.IX was based on the potential challenges and difficulties students faced trying to be interactive in the class.
The emerged data reveal that in most of the EFL classes, the interaction pattern followed is T-Ss and the majority of the student respondents experienced a poor show of student-student interaction inside and outside the EFL classroom. In addition, it is also found that the percentage of truly interactive classes was very poor. This implies that the tertiary EFL classes were largely teacher-centered which is also supported by the findings of Mridha and Muniruzzaman (2020). They find that our tertiary EFL classes were mostly teacher-dominated and an ideal interactive scenario is largely absent. Regarding teacher-initiated interactions, the majority of the students also felt that the rate of teacher-student communication in the EFL classroom was not satisfactory enough. This finding is highly parallel to that of Mridha and Muniruzzaman (2020) who report that only "6.66% often and 0.95% respondents always speak English" (p. 125). The results also go against the suggestion of the scholars regarding the expected amount of student-student interactions to take place in the EFL classes. Tomlinson and Dat (2004) suggest that EFL students will be able to speak better if their teachers encourage oral participation, nurture a supportive atmosphere, and encourage peer interaction. Although it is seen that in the Tunisian context this support is satisfactory enough in the tertiary EFL classroom (Ounis, 2017) this picture is quite different in the Bangladeshi context.
With regards to students' satisfaction level towards the efficacy of the classroom interaction, it is seen that the majority of the students were not happy with the outcome of the classroom interaction and the vast majority of the students believed that most of the EFL classes were not truly interactive. This implies that the Bangladeshi tertiary EFL classrooms are still not practically ready to develop students' speaking abilities to the expected level. The research results further demonstrate that a good number of students did not find the lessons interesting and the classroom teaching/learning environment supportive enough to enhance their participation in the interactive activities. This finding further supports the work of Farooqui (2007) who observes that students come to the tertiary level in Bangladesh with an educational background where teachers just delivered the syllabus and organized the exams "leaving little room for genuine enquiry, questioning or criticism" (Thornton, 2006).
With regards to the teachers' address of speaking errors, the majority of the students opined that their teachers were positively active in correcting their mistakes and errors and giving feedback in speaking lessons. It is also worth observing that a good number of the teachers were not found active in this role. In this regard, Murray and Christison (2011), Ur (2009), and Walsh (2011 opine that in the classroom, traditionally controls are at the teacher's hand as the knower and they should give feedback or correct the speaking mistakes when students speak. Furthermore, the learning environment requires teachers to play various roles to fulfill students' different needs but the data show that teachers mainly play two major roles: the role of a corrector and a prompter. This implies that most of the teachers mainly focused more on mistakes than on the flow of interaction which is not a helpful practice for developing learners' speaking fluency as their practice is interrupted very frequently and they may feel low for their mistakes. Teachers will correct the mistakes made but their main target should be allowing the pupils to produce language without interruption and fear (Brown, 2007).
During interactive sessions in the EFL classroom, the data reveal that the students encountered many obstacles. A good majority of them (86%) lacked enough language resources and confidence to speak and interact in English. This finding is very synonymous with those of Farooqui (2007) who reports that "students usually have a small English vocabulary" (p. 102). For this reason, they lack confidence, feel shy, and do not want to speak in front of the class (Farooqui, 2007). Farooqui (2007) also observes that though university EFL teachers try their best to create a favorable environment, most of the students feel anxious and lack the courage to initiate any sort of conversation.
Teachers' interview sheets also contained 11 questions to collect their perceptions about the classroom interaction practices in the EFL classes they taught and the challenges they faced in developing students' interactive speaking skills.
The findings from the teachers' interviews reveal that in the Bangladeshi tertiary EFL classroom, interaction included not only some oral practices but also personal and pedagogical practices. Some of the teachers said they applied S-S, T-S, S-Ss patterns because they thought that their students' speaking skills would develop further if they had talked with their classmates more than with the teachers themselves. It is also seen that in the EFL classroom teachers mainly used English to manage the class but more often than not they used both Bangla and English in their lectures. They said they had to use both the languages in order to build rapport with their students and explain the lessons better. The data also demonstrate that teachers put effort to create a congenial atmosphere in the EFL class and involved students in interactive speaking skill activities. This might prove to be consistent with the second language classroom activities suggested by Harmer (2011) and second language classroom modes by Walsh (2011). Undoubtedly, this managerial mode is one of the primary classroom interaction features which are used to organize a successful physical classroom environment. It is also demonstrated by the data that students tend to learn much better if they work in pairs or groups because this shared activity develops their intellectual understanding, communication skills, and managerial skills; eradicates their gaps of knowledge; and helps them become more fluent in 242 CROSSINGS: VOL. 12 | 2021 | ISSN 2071-1107 Classroom Interaction as a Way of Developing Students' Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation time. This view is also supported by Khan (2017) who observes that students learn better in group work.
Regarding the types of interactive activities teachers mostly used in the speaking classes, the data demonstrate that most of them organized group work, pair work, role plays, warm-up sessions, using multimedia to facilitate and motivate students to interact in the classroom. The finding agrees with the suggestions of Harmer (2011) who confirms that a language teacher should organize different activities like pair work, group work, role-plays, and different spoken discourse and interactive activities in the EFL classroom.
The data collected from the teachers' interviews also demonstrate that they face some realities and practical challenges which hindered their classes from being truly interactive. The major ones were students' poor vocabulary stock and limited linguistic materials, gaps among their language proficiency levels, lack of proper academic environment and motivation, a poor EFL background, and a very limited time allocation for interactive activities in the EFL classes. Therefore, it can be suggested that institutional policies, curriculum, students' exposure to the language, and their social and educational background are the factors that seemingly influence how teachers and students interact with each other. This research finding may also support the suggestions offered by Stern (1983, in Hall, 2011 and Seedhouse and Jenks (2015) about different contextual factors that are involved in language teaching. They suggest that the culture of the classroom interaction might also be affected by the institutional/national policies and learners' linguistic or cultural backgrounds. They also suggest that classroom interaction can be affected by several variables outside the classroom too.
To sum up, the results from this study confirm both the positive influence and helpful effects of classroom interaction on students' speaking skills. This research exposes some ground realities and factors that affect the development of the interactional speaking skills of tertiary EFL students of Bangladesh. Although both students and teachers showed positive attitudes towards an interactive and engaging classroom environment, it is evident that Bangladeshi tertiary EFL classrooms are still not prepared to deliver the services to a satisfactory level.

Limitations and Recommendations for Further Research
Though this paper has produced a good number of findings regarding the classroom interaction scenario and tertiary level students' speaking skills development in Bangladesh, it has some limitations too in terms of samples, instruments, and task types. First, it offers findings from the respondents that represent only one region of Bangladesh and this suggests that the data is very limited to be considered a generalized finding. Second, only two instruments were used to collect the data: a questionnaire survey for students and semi-structured interviews for teachers. Data triangulation and methodological triangulation were not largely employed. Third, only one type of tertiary institution, namely the private university was included in the study (four of them). If the samples had been selected from some more regions and they had represented all types of universities and tertiary institutions, and if proper data triangulation and methodological triangulations had been employed, the study might have generated more generalized and comprehensive data. For future research, this study suggests that the researchers include respondents from all types of tertiary institutions representing all regions of Bangladesh. It also suggests that future researchers should apply more data collection methods and include more data sources. All these steps, if implemented, will help us develop a richer database and explore more solid findings in the field of interactional speaking skill development.

Conclusion
This research highlights the importance of classroom interaction for the overall improvement of the Bangladeshi tertiary EFL students' speaking skills and the findings strongly suggest that classroom interaction can be a very effective tool to develop their communicative abilities. Consequently, extensive classroom interaction is highly recommended in the tertiary level EFL classroom in Bangladesh as it fosters an exchange of knowledge, expedites autonomous learning and builds up confidence in the students, consequently promoting their expertise in oral communication.
The results also assert that although both students and teachers strongly believe that classroom interaction is a very effective strategy to develop students' speaking and communication skills, Bangladeshi tertiary level EFL classrooms are not still prepared to implement the effective practices of classroom interaction techniques in the truest sense and consequently being deprived of its potentials to offer deserving outputs to the tertiary learners.
In the light of these findings, the summary of this entire work can be illustrated in the following two main points: Bangladeshi tertiary students want to see their ELT teachers in the role of better organizers, prompters and resource persons. Teachers with these three key roles will then assist them to interact and speak better in the EFL classroom and make them confident users of English.
Teachers should incorporate cooperative learning techniques as a mandatory teaching-learning activity in the tertiary EFL classroom where learners can think and research independently and interact in pairs or groups comfortably. Teachers also need to provide students with more opportunities in the 244 CROSSINGS: VOL. 12 | 2021 | ISSN 2071-1107 Classroom Interaction as a Way of Developing Students' Speaking Skill at the Tertiary Level EFL Classroom: An Empirical Investigation classroom to interact with each other since they enjoy group work and pair work the most.
Above all, this research elicits some insights into the relationship between classroom interaction and students' speaking proficiency. These insights could be useful to the tertiary EFL teachers in being informed of the ground reality more and strengthening their classroom teaching strategies. These could also help the tertiary EFL students develop their understanding of a pragmatic way of developing their speaking skills.