Luận án Nghiên cứu nhận thức của giáo viên và sinh viên về các yếu tố ảnh hưởng đến việc nói tiếng anh lưu loát ở trường Đại học tại Việt Nam

The following factor is exposure to English (question numbers 38, 39, 40, 41, and 42). The final factor theme is the technology theme (question numbers 43, 44, 45, 46, and 47). Table 3.4. Summary of the Questionnaire Instrument with closed-ended items for quantitative data Themes Items 1. The importance of LOF 1, 2, 3, 4, 5, 6, 7 2. Student Behavior And Attitude Factors 8, 9, 10, 11, 12 3. Motivational Factors 13, 14,15, 16, 17 4. Anxiety Factors 18, 19, 20, 21, 22 5. Task Type Factors 23, 24, 25, 26, 27 6. Instruction Factors 28, 29, 30, 31, 32 7. Environment – Class Size Factors 33, 34, 35, 36, 37 8. Exposure Factors 38, 39, 40, 41, 42 9. Technology Factors 43, 44,45, 46, 47 3.4.2. Semi-structured interviews for teachers and students To consolidate the findings and triangulate the data collection method, the researcher also conducted an interview with the participants. Semi-structured interviews were employed as the data collection method. Qualitative researchers commonly use interviews to discover and portray the multiple views or realities of a case (Stake, 1995). In other words, it is a valuable way to see how people interpret the world around them (Merriam, 1998). Moreover, qualitative interviews are a typical way to collect data and are generally depicted as a discussion (Kvale & Brinkmann, 2009). "Interviewing is crucial when we cannot witness behavior, feelings, or how people understand the environment around them," Merriam (1998, p. 72) remarked. Therefore, interviews are appropriate and applicable for this study because it is better to know what Vietnamese teachers and students understand and interpret the concept and effects of LOF. Three types of interviews are widely used concluding structured, semi-structured, and unstructured (Hancock & Algozzine, 2006; Johnson & Christensen, 2012; Merriam, 75 1998). The first method is used to collect general socio-demographic information from respondents. The second method allows the researcher to deal with situations, the respondent's emerging worldview, and new ideas on the particular subject. The final method is rarely used as a sole method because the researcher may not know enough about a phenomenon to ask crucial questions during the interview. The researcher used semi-structured interviews to achieve the particular purpose of this research. Semi-structured interviews were employed in this study to "understand more thoroughly" (Kuş, 2009, 87) EFL teachers' and students’ perspectives on the factors impacting speaking fluency. In other words, the interviews are semi-structured, with leading questions that allow us to dig deeper or follow up as needed to extract more information from the participants' responses (Alvesson, 2011). A semi-structured interview design with six open-ended questions was created for this aim. Two field educators from a state university's Department of Foreign Languages were consulted to assess the content validity of the questions. One specialized in educational assessment and evaluation, while the other specialized in teacher professional development. We attempted to create a safe environment for the interviews. The participants were told that their identities would be kept private. More importantly, their consent was obtained before the recording of the interviews. Interviews were carried out by phoning participants on the Zalo social network, which took about 25– 30 minutes. This was due to the fact that the Zalo social network is widely used and convenient in Vietnam and the COVID outbreak in cities in central Vietnam. Certain factors are thought to be considered when the researcher sets up and conducts an interview. For example, the social interaction between interviewers and interviewees may result in bias (Wellington, 2000). Unfortunately, due to the COVID- 19 pandemic, the researcher had to arrange the interview through Viber messengers for the current research. According to Reosa, M., Mwamba, C., Meghani, A., West, N. S., Hariyani, S., Ddaaki, W., Sharma, A., Beres, L. K., & McMahon, S. (2021), "we champion the value and necessity of rapport building, empathy, open and honest dialogue, and a sense of closeness between research teams and interview respondents." However, they also suggested that using phone interviews to gather qualitative data resulted in a lack of nonverbal data, contributing to limited knowledge of context. In this research, interviewing via telephone is a must in this situation. On the other hand, 76 several others emphasized the advantages of phone interviews in providing richer discussions on sensitive topics due to increased perceptions of privacy and confidentiality. Furthermore, telephone interviews improved access to difficult-to- reach respondents and settings that may be considered unsafe for research. Given the focus of the current study in terms of observing the change in teachers' and students' understanding, the interview was aimed at comparing the teachers' and students' views on the questionnaire phrase and the interviewing phase of the study. A qualitative comparison of each teacher's and students' responses to the questionnaires allowed the researcher to identify different degrees of change in the way different teachers and students defined and understood fluency. This comparison also demonstrated variability among the participants in the extent to which they drew on the new concepts and definitions understood by them. Moreover, the interviewing method is used to examine the factors of LOF from EFL teachers at the universities in Vietnam when implementing the appropriate teaching methods to meet the fluency learner outcome. Finally, this approach also improves the evaluation of major English students toward the appropriate teaching and learning methods of EFL teachers and the capacity to meet the course's fluency learning outcomes and that of the school's curriculum. We will use the results to make the interview survey better and clarify the status in accordance with the objectives of the research study. The current article is heavily based on quantitative analysis data. Although the findings of this qualitative study cannot be directly generalized to other contexts, a detailed description of the findings may provide readers with an experience that they can then generalize appropriately to their own experience (Stake, 1978). According to Stake (1995, p. 37), qualitative researchers have "pressed for explanation and control" and "understanding the complex interrelationships among all that exists." More than that, the sources gathered through qualitative data verify or reject results from quantitative data. Table 3.5 shows the key questions that were raised during the interview. Table 3.6 shows the teachers who took part in the interviews and the questions they were asked, whereas table 3.7 shows the students who took part in the interviews and the questions they were asked. 77 Table 3.5. Interview questions for teachers and students 1. How often do you/ your students communicate in English in class and outside of class? 2. What does the term "oral fluency" mean to you? 3. Do you think speaking fluently is important for you/students when learning a foreign language? 4. Of the following factors, which influence most students' oral English fluency? You can choose up to 3 factors and explain why. a. Behavior and attitude factors b. Motivational factors c. Anxiety Factors d. Task-Type Factors e. Instruction Factors f. Environment: Class Size Factors g. Exposure Factor h. Technology Factors 5. Of the following factors, which influence the least amount of students’ oral English fluency? You can choose up to 3 factors and explain why. a. Behavior and attitude factors b. Motivational factors c. Anxiety Factors d. Task-Type Factors e. Instruction Factors f. Environment: Class Size Factors g. Exposure Factor h. Technology Factors 6. Do you think using technology can help improve oral speaking? If so, why? If not, why not? Table 3.6. Coding for interviewing teacher participants Participants Identification numbers Interview dates Teacher 1 T1 09 May 2021 Teacher 2 T2 11 May 2021 Teacher 3 T3 14 May 2021 Teacher 4 T4 18 May 2021 Teacher 5 T5 18 May 2021 Teacher 6 T6 18 May 2021 Teacher 7 T7 21 May 2021 Teacher 8 T8 21 May 2021 Teacher 9 T9 21 May 2021 Teacher10 T10 26 May 2021 Teacher 11 T11 26 May 2021 Teacher 12 T12 26 May 2021 78 Table 3.7. Coding for interviewing student participants Participants Identification numbers Interview dates Student 1 S1 07 May 2021 Student 2 S2 07 May 2021 Student 3 S3 12 May 2021 Student 4 S4 12 May 2021 Student 5 S5 21 May 2021 Student 6 S6 24 May 2021 Student 7 S7 24 May 2021 Student 8 S8 24 May 2021 Student 9 S9 25 May 2021 Student 10 S10 25 May 2021 Student 11 S11 25 May 2021 Student 12 S12 26 May 2021 3.5. Data collection procedure The information was gathered from primary and secondary sources. The primary data sources were questionnaires distributed to students and teacher educators. Documents collected while visiting each site served as secondary data sources. These documents were among curriculum guidelines, course texts, and conference and seminar proceedings. The researcher is the primary data collector and analyst (Creswell, 2012). For this current study, the method of qualitative research and quantitative surveys are used with objects during the school period. In other words, two instruments were employed to collect data in numeric and non-numeric forms for this study. The procedures for data collection consisted of piloting and administering the instruments. 3.5.1. Piloting the questionnaires According to Cohen (2007, p.341), a pilot has several functions and is described as a way "principally to increase the reliability, validity, and practicability of the questionnaire." A pilot study is also a small-scale trial of research that is conducted before a full-scale experiment (Gay et al., 2011). This means that piloting is done primarily to verify that the research instruments can be used successfully in the actual study. According to Bryman (2012), "piloting makes it simple to identify elements that 79 are ambiguous or unlikely to be of importance in the topic because they do not form a variable." Thus, piloting the instruments is critical for ensuring content validity (Creswell, 2014). For the present study, the questionnaire, which included 47 open- ended items, was piloted with 12 EFL students and 12 EFL teachers who were similar to the planned population of our study. In this research, the first step was to collect survey research to evaluate the reliability and validity of this questionnaire instrument. Before being applied, the "LOF" questionnaire was thoroughly prepared and translated into Vietnamese. Before piloting the questionnaire, two qualified and experienced professors were asked to provide feedback on their understanding of the question items to prevent the drawbacks of the instruments. The recommendations of these two experts in terms of word choice, format, and grammatical structures were taken into consideration. Finally, during the piloting stage, the researcher submitted the English questionnaire versions to 12 EFL students and 12 EFL teachers by using an online survey on Google Form. The data from the piloted questionnaire was then encoded and analyzed using IBM SPSS Statistics 18. Suppose the Cronbach's alpha (α) value is 0.7 or higher. In that case, it indicates that the piloted questionnaire is a reliable tool that can be used to collect data for actual research (Nunnally and Bernstein, 1994, p. 265). The results for the Cronbach’s Alpha (as shown in tables 3.8 and 3.9) revealed that the internal consistency of the entire questionnaire was 0.885 for teachers and 0.721 for students, indicating an acceptable and high level of reliability for items in the instrument. Hence, the items of the questionnaire conducted in the research were validated and were reliable for data collection in the next phase. Table 3.8. The reliability of teachers’ pilot questionnaire Administering the questionnaire Number of teacher participants Reliability (Cronbach’s Alpha Coefficient) Pilot questionnaire 12 0.885 Table 3.9. The reliability of students’ pilot questionnaire Administering the questionnaire Number of student participants Reliability (Cronbach’s Alpha Coefficient) Pilot questionnaire 12 0.721 80 3.5.2. Piloting the semi-structured interview The researcher's supervisor proofread and revised the semi-structured questions before they were officially used. The researcher conducted interviews with EFL teachers and students who were chosen from the samples based on their willingness to participate in order to acquire a better understanding of their understanding through their responses. The researcher contacted the teacher and student participants in the questionnaire survey by email, Zalo social network, messenger, or phone calls after they were selected on the basis of their voluntariness. The responders were given advance notice of the time, location, and form of communication. Interviews were conducted in English to ensure the dependability and comprehensibility of questions, as well as to provide interviewees confidence in providing as much information as possible. Finally, the researcher was able to reach into their ways of teaching in their class in relation to the second study topic by asking meaningful questions because the privacy issues associated with each teacher's teaching methods are quite sensitive. In summary, three pilot interviews enabled the researcher to develop the interview instrument, allowing her to collect a huge amount of qualitative data for the research findings. 3.5.3. Administering questionnaires Firstly, the questionnaire's reliability and validity were tested using SPSS version 18.0. It suggests that the Statistical Package for the Social Sciences (SPSS) version 18.0 was used to examine the quantitative data collected from the questionnaire. Following the pilot, a 47-item questionnaire was distributed to gather information on EFL teachers' and students' perspectives on the importance of LOF for EFL students and potential factors contributing to the LOF. After that, a scale test was used to determine the whole questionnaire's dependability. The questionnaire's reliability coefficient (as stated in Table 3.10 and 3.11) is high (=.868) for teacher participants and (=.822) for student participants, according to the results. As a result, the items in the research questionnaire were validated and were reliable for data collection. Table 3.10. The reliability of the official questionnaire- Teachers Administering the questionnaire Number of participants Reliability (Cronbach’s Alpha Coefficient) Official questionnaire 45 .868 81 Table 3.11. The reliability of the official questionnaire- Students Administering the questionnaire Number of participants Reliability (Cronbach’s Alpha Coefficient) Official questionnaire 115 .822 Then, scale tests were also conducted to assess the reliability of teacher and student questionnaires as themes, typically the reliability coefficient of nine themes (as shown in Table 3.12), including the reliability coefficient of the importance of LOF (the cluster of theme 1 of the official questionnaire), the reliability coefficient of student behavior and attitude factors theme (the cluster of theme 2 of the official questionnaire), the reliability coefficient of anxiety factors theme (the cluster of theme 4 of the official questionnaire), that of task type factors theme (the cluster of theme 6 of the official questionnaire), that of Environment—class size factors theme (the cluster of theme 7 of the official questionnaire), that of exposure factors theme (the cluster of theme 8 of the official questionnaire), and that of technology factors theme (the cluster of theme 9 of the official questionnaire). Table 3.12. The reliability of the theme - Teachers Administering the questionnaire Number of participants Reliability (Cronbach’s Alpha Coefficient) Theme 1 45 .758 Theme 2 45 .652 Theme 3 45 .719 Theme 4 45 .842 Theme 5 45 .784 Theme 6 45 .784 Theme 7 45 .668 Theme 8 45 .660 Theme 9 45 .901 As can be seen from Table 3.12, the reliability of nine themes of teacher questionnaires were acceptable (α =.758, α =.652, α =.719, α =.842, α =.784, α =.784, α =.668, α =.660and α =.901). Besides, Table 3.13 showed the Cronbach alphas for those of students were above.65 and ranged from.655 to.860. Consequently, these tests were sufficiently reliable to be applied as instruments for the current study. 82 Table 3.13. The reliability of the theme - Students Administering the questionnaire Number of participants Reliability (Cronbach’s Alpha Coefficient) Theme 1 115 .813 Theme 2 115 .661 Theme 3 115 .661 Theme 4 115 .860 Theme 5 115 .754 Theme 6 115 .739 Theme 7 115 .813 Theme 8 115 .655 Theme 9 115 .720 3.5.4. Administering the semi-structured interview To guarantee the reliability and validity of the interview instrument, the researcher conducted interviews with three EFL teachers and three EFL students who were selected from the samples based on their willingness to participate in order to acquire a better understanding of their comprehension through their responses. The researcher was able to improve the interview instrument attributed to the following pilot interviews. First, the researcher gained an understanding of the perspectives of the participants on the research topic. Next, the researcher looked into the difficulty level of the interview questions, changing them to make them simpler in order to fully exploit the information for research topics." what are the key factors influencing students' oral fluency?" the researcher, for example, used to ask. The interviewees could not give you proper answers in such a limited amount of time, and this question is tricky. The researcher then tried to elicit what they think about the factors, such as "Of the following factors, which influence the most students' oral English fluency? You can choose up to 3 factors and explain why?". In short, three pilot interviews allowed the researcher to develop the interview instrument, allowing her to collect a substantial amount of qualitative data for the research findings. Before they were formally used, the semi-structured questions with interviewing suggestions were sent to the researcher's supervisor for proofreading and revision. More than that, the researcher revised her interview questions with the guidance of her supervisor and another professor. 83 The researcher then proceeded to conduct nine interviews, each of which lasted 20 to 25 minutes and was meticulously recorded by the researcher with the participants' agreement. Following that, the original English transcripts of the interviews were meticulously transcribed. The contents of the interviewees' English transcriptions were then evaluated for topics and themes related to the research goals. Semi-structured interviews (see Appendix 3) with main questions and sub-questions were conducted one-on-one and over the phone to discover more about participants' perspectives on the importance of "LOF" for university students, as well as EFL teachers' and students’ perspectives on factors influencing "LOF." 3.6. Data analysis procedures As there were two types of instruments used to collect the data for this present study: questionnaires and interviews, the data would be analyzed qualitatively and quantitatively. The study of field data as it is acquired in order to detect emergent themes is a feature of qualitative research. The results of qualitative data analysis can be analyzed in various ways (Merriam, 2009). In addition to the qualitative data acquired, quantitative data can give insight, support, and context (Saldana, 2011). In order to analyze and interpret the data obtained from the questionnaire, the data will be entered and managed via the software Epidata 3.1 and processed and statistically analyzed via SPSS 18.0. The researchers will analyze the material acquired during the interview by using the general steps of the theme analysis approach. According to the definition, "a strategy for detecting, evaluating, and reporting themes within data", it organizes and defines your data set in great depth, at a minimum (Braun & Clarke, 2006, p. 79). The current study's descriptive survey was done in the second semester of 2020– 2021. The questionnaire was distributed to students in speaking classes during the second semester. In addition, teacher and student interviews were conducted in order to obtain more detailed information for data analysis. The questionnaire was given to students in speaking classes in the second semester of the academic year to get a holistic view of the role of technology in providing opportunities for students to improve their English speaking fluency. In the second semester of the school year, students were given a 47- item questionnaire, including parts A and B. Online survey questionnaires were created by using Google Forms, and their links were sent via email to 120 English-speaking respondents and 45 teachers. The teachers of the class involved in the survey explained that students' anonymity would be kept safe and confidential. Also, they helped explain all 84 items thoroughly to ensure that all students understood before making a choice and also to ensure a 100% response rate. The respondents were given 20 minutes to complete the questionnaire. The online survey questionnaires were then collected by downloading the excel file. The survey was completed with the support and collaboration of the teacher, who was responsible for the selected speaking classes successfully. The survey was completed and recorded in an SPSS version 18 datasheet for further analysis to determine the current state of the impact of factors on boosting speaking fluency in the Faculty of Foreign Languages at universities in Vietnam. Following that, the collected survey data were analyzed using SPSS to calculate the means and standard deviation, total number of participant responses, frequency, and percentage marking on each response for each statement. This quantitative data was organized into nine clusters in order to be combined with qualitative data. The second phase of data analysis was the translation and transcription of interview data On the other hand, semi-structured interviews were the second study approach used to gather data. Twelve full-time teachers and twelve students participated in the interview, which lasted around 20 minutes for each. Due to the COVID-19 epidemic, the participants were instructed to answer the questions through Google meet or Zalo. At first, these researchers described the goal of the interviews so they knew what they were doing. The interviews were recorded and transcribed, and the transcriptions were used in the qualitative analysis of the study. The qualitative data acquired during the interview is analyzed using three techniques. First, listen to the interview via online recording and note down ideas for each question; second, read all the statements and categorize each interviewee's responses; third, check over all of the responses for each topic. To summarize, the researcher has thoroughly examined the data collected through questionnaires and interviews with respondents to determine its importance in light of the study's objectives. Also, the online survey questionnaire and the interview data were analyzed separately before the findings were brought together to answer the research question. 3.7. Reliability and Validity Regarding the concepts of generalizability, validity, and reliability, this thesis is associated with Cohen et al.'s (2011) and Oppenheim's (1992) notions; in addition, other perspectives on validity and reliability in mixed method research will be discussed in relation to the learning studies conducted. According to Cohen et al. (2011), internal validity means that the findings must accurately describe the phenomenon under investigation. For example, the reliability of attitude statements, such as those in the questionnaire, was 85 attempted u