An Exploration of the Factors Hindering Students’ Lesson Comprehension in EMI Classes

The implementation of English as a medium of instruction (EMI) has been found to bring numerous benefits to its learners. There has been a rising trend in introducing EMI at higher education institutions worldwide. However, challenges are ubiquitous and there is still prevalent hesitation in integrating content and language in many institutions. This study aimed to explore factors hindering students’ level of lesson comprehension in EMI classes through a mixed-method study carried out on 233 participants at Hoa Sen University, Vietnam. The findings showed that variables hampering students from lesson understanding were mainly associated with instructors’ teaching methods and students’ and instructors’ language competences. Results from standard multiple regression indicated that vocabulary range, writing skills and teaching methods were found to have significant contribution to the prediction of the level of lesson comprehension of the students. The findings of the study could be a considerable reference source for universities aiming at implementing EMI programmes. Received: 31/05/2020 Accepted: 07/09/2020

conducted by Hellekjaer (2010) on a number of Norway and German students discovered that the students found it difficult to read texts or take notes in EMI classes because of unknown words. It was argued by Coxhead and Boutorwick (2018) that without the suitable input from reading students have insufficient knowledge of words to deal with subject content. Hence, students tend to get lost in lecturers' speech and cannot take note by their own and wait for lecturers' writing key words or technical terms on board to copy (Chuang, 2015). In the context of Vietnam, limited vocabulary knowledge was found to be a greatest source of difficulty preventing students from not only understanding the knowledge content but also expressing their opinions comprehensively EMI classes .

Inadequate Language Skills
Students' inadequate language skills are also found to be among those of the great barriers in EMI classes. In fact, using English proficiency test and subject exam scores in their research, Rose, Curle, Aizawa and Thompson (2019) asserted that low English proficiency significantly influenced the success of EMI programs. According to Keuk and Tith (2013), reading skills seem the most difficult tasks for students because of grammatical differences between English and students' mother tongues. Besides, students encounter uncounted challenges in writing assignment when writing in a language not of their mother tongues (Al-Kahtany, Faruk & Al Zumor, 2016;Rogier, 2012). Students' essays lack both diversified uses of vocabulary and critical thinking (Dorasamy, 2012). Keuk and Tith (2013) believed that it is the lack of accurate input and regular practice resulting in weak writing skills of students. Kagwesage (2013) found that students in EMI classes have some difficulties in communication due to the lack of language skills. They can only produce single and simple word expressions when getting involved into classroom discussions. Additionally, learners are afraid of making mistakes when communicating in the language not of their mother tongues. This gradually make them feel inhibited and not confident in their speaking (Keuk & Tith, 2013). Consequently, students seldom ask and answer questions, expand issues in discussions, and well-communicate in classes. Joe and Lee (2012) indicated that poor speaking ability prevents students from interacting with lecturers, so students often use mother tongue to ask and discuss in group.
Listening is a mean for learners to absorb knowledge during class time. It strongly influences the extent of lecture understanding of the learners especially in EMI classes. However, many students admit that they do not have good listening skills and this negatively influences their lesson understanding and course achievement (Kagwesage, 2013;. Likewise, a research of Alanshory (2014) indicated that 77% students admit that listening skills hinder course acquisition and that 98% EMI instructors believe students have trouble comprehending academic content in English due to inability of English listening.

Teachers' Oral Skills
Factors hindering students from understanding the lessons do not merely come from the students themselves. Hu (2019) concluded upon reviewing four empirical studies of EMI in Chinese universities that instructors need to be adequately capable of the EMI. Without it, instructors are not likely to -engage students in complex cognitive processes, scaffold their effort to master disciplinary knowledge or provide rich language input to develop advanced English proficiency in the students‖ Hu (2019, p.8). A couple of factors have been found coming from the lecturers that make students struggle to understand the lessons. These are lecturers' oral skills and their teaching methods (Civan & Coskun, 2016;Dorasamy, 2012).
Lecturers' poor oral skills are considered to be the main cause for the loss of students' understanding (Chang 2010;Civan & Coskun 2016). According to Bradford (2013), over 80% of students indicates that lecturers' speaking skills are insufficient. Inaccurate pronunciation and shortage of vocabulary are two common problems. The limited vocabulary range makes lecturers have difficulties in expressing effectively, so they rarely paraphrase, explain clearly, and refine statements (Vu & Burns, 2014). Another factor coming from lecturers making students feel challenging to keep up with instructor's lecture is their fast speed (Goodman, 2014;Kagwesage, 2013). Likewise, Vu & Burns (2014) pointed out that lecturers' speed affects much not only students' comprehension but the learning attitudes as well. Sometimes, lecturers often feel anxious even though their English competences are more than enough to teach EMI subjects (Goodman, 2014).

Teachers' Teaching Methods
Teaching methods have found to make a significant contribution to students' lesson comprehension (Kim, Kweon & Kim, 2017). Bradford (2019) even asserted that teaching methods, such as teaching practices which focus on interactive discussions, are considered more important for lesson comprehension than instructors' imperfect English ability or student listening skills. He argued for the necessity to pay greater attention to pedagogical measures for the success of EMI programs. Vu and Burn (2014) believed that lecturers are offered few opportunities of appropriate language training before taking EMI classes and they argued for the necessity of EMI pedagogy in order to avoid unclear guidance in EMI lessons. Specifically, code-switching method in EMI remains controversial. Some authors hold the view that code-switching is necessary for explaining difficult concept or challenging content because of exclusive use of English brought to difficulties in comprehension (Hu & Lei 2014;Al-Kahtany et al. 2016). However, other researchers doubted that code switching may not be an appropriate pedagogical approach as it fails to facilitate learners' academic literacies (Kyeyune, 2010).

Time Constraint
Time limitation was found to play as a significant hindrance affecting the success of EMI classes in the research of Keuk and Tith (2013). It has been argued that EMI classes offer an inadequate duration for lecturers and students to work out the lessons. Similar finding was also found in the case of Dutch students. Learners admitted that they could not comprehend a large amount of information delivered in the class in a limited amount of time and respondents also claimed that more time should be allocated for the learners to finish class assignments (Wilkinson, 2005). Additionally, due to limited knowledge of both academic and specialized vocabulary, a large proportion of students spend extra time to read and understand textbooks and materials with the support of dictionaries or translation devices .

Methods Applied
Mixed method was selected for this study because its benefits. According to Hesse-Biber (2010), this combination of quantitative and qualitative approaches can take advantages of strengths limit the disadvantages of each individual approach. Particularly, quantitative leads researchers to conclusion by analyzing figures and statistics which means the method allows the researchers to investigate a phenomenon with a wide range of research facets or items (Neuman, 2007). The survey gives researchers a chance to approach a great amount of data that need collecting and analyzing (Hesse-Biber, 2010). Qualitative research approach, on the other hand, focuses on gaining a various range of responses, which provides profound understanding of research issues through participants' views and thoughts (Berg, 2001;Creswell, 2012). The interview is implemented to find out insightful information which provides researchers a full explanation of what, why, where, and how specific phenomenon occurs (Neuman, 2007).

Instrument Construction
The questionnaire was designed after analyzing literature and reviewing the major focus of the study. The questionnaire consisted of five sections. This included personal information, difficulties students encountered, benefits students gained, students' satisfaction, and three open ended questions for students to express their perceptions. More specifically, there are 40 questioned items in total related to of EMI classes, including two main types: rating scale and multiple choice. These closed-ended questions help to collect and code the basic background information of participants. For the scale of this paper, only 18 items of the first two sections related to the demographic background of the research participants and the difficulties students encountered in their EMI classes were selected for analysis.
All the items in the questionnaire were constructed by referring to the studies of previous research suitable for this study context. The content of the questionnaire was edited and double-checked using the content validity techniques by discussing with four university instructors who are knowledgeable of EMI and a group of six students from different majors and academic years who were taking EMI classes at the time of getting feedbacks. From the feedbacks of the instructors, students, the questionnaire was clarified in other that no ambiguity and misunderstanding remained. As a result, the final version of questionnaire was accomplished by considering whether questions covered all main aspects that the study aims to explore. Internal consistency reliability was employed in this study to check the consistency among the variables. This would help to see whether all the scale measures the same underlying attribute or not. The Cronbach alpha coefficient of 10 items concerning difficulties students encountered in their EMI classes used in this paper was .833. The figure proves that the scales have fairly good internal consistency.

DATA COLLECTION
Data collection was conducted via pencil-and-paper interviews, face-to-face interviews and group discussion. At the time of data collection, three faculties namely Economics and Business, Science and Engineering, and Tourism of Hoa Sen University offered EMI classes. Cluster sampling technique was used to pick up two EMI classes each faculty and all the students were approached for survey questionnaire. Cluster sampling was selected because it helps to pick up the respondents relatively equal from all three faculties. 300 questionnaires were printed out and distributed to the participants during their class time in May 2018. A cover letter was attached to each questionnaire to indicate the purposes of the study and ask for the consent of the participants in filling out the questionnaire. Furthermore, promises were also made to the participants that their information would be kept confidential under any circumstances and that the informants had their own rights whether to join in the study or not. A total of 282 (94%) papers were returned and 233 (77.6%) responses were used for analysis.
A total of 20 students from three faculties were also invited for group or individual interviews which were recorded or note-taken to reinforce the findings of the survey. The interviews were conducted in Vietnamese. Although all the informants had at least basic command of English when taking the EMI classes, interviews conducted in their native language would make them easier to discuss the issue of EMI implementation. This is also to make sure that no barrier would happen in terms of language to prevent the informants in expressing their viewpoints. Purposeful sampling technique was used to select the interviewed informants. The criteria for informants invited for the individual and group interviews were those who had experiences in taking EMI classes, had comprehensive understanding of EMI and had the willingness to share their viewpoints on the issues of EMI implementation. The informants invited for the interviews came from all the three faculties. This not only helped to diversify viewpoints but also ensured the balance among different students of different year levels and majors. The focuses of the interviews were to explore the informants' personal viewpoints, experiences and feelings towards the difficulties that they faced, the benefits that they believed that they gained in the EMI classes. The table below shows the general background of participants of this study including gender, student year, faculty, and the number of EMI courses taken at school.

DATA ANALYSIS
SPSS version 22.0 was utilized to analyze data collected from the survey questionnaires. Particularly, descriptive analysis was employed to count some variables such as gender, year of learning, faculty, the number of EMI courses taken, and the frequencies of difficulties. Moreover, multiple regression was used to discover the correlation between challenges students faced in EMI courses and the overall EMI lesson comprehension. This also helped to evaluate how well items of challenges in taking EMI courses could significantly predict the research participants' lesson comprehension.
Qualitative data analysis was analysed based on the steps of data analysis suggested by (Creswell, 2012). Firstly, audio recordings and fieldnotes were transcribed. Data was then closely examined for the general sense. Manually open coding technique was employed. Data was next checked for similarity and differences, and then categorized under themes. Theme labels were generated with the reference of the themes obtained from the review of literature. The items that relevant to the factors hindering students' lesson comprehension in EMI classes were selected for analysis in this paper. Descriptive analysis of lesson comprehension in EMI classes shows that among 232 research participants gave the responses to the questions about the degree they could comprehend lessons taught in English (10-point scale). Only 1.7% indicated that they totally understand the lessons (100%). 83.2% percent of the respondents mentioned that they understand the lessons in EMI classes 50% or more, with the highest figure lies in those who understand 70% of the lesson (23.7%). 16.8% showed that they faced lots of difficulties in lesson comprehension (40% or less).

Lesson Comprehension in EMI Classes at Hoa Sen University
The findings were almost in line with those of a study conducted on 370 undergraduate students in Taiwan by Chang (2010). Chang found that very few (5.8%) students are able to understand completely the lectures in English and only 24% students think that they understand approximately 75 percent and around 40% of the students comprehend the degree of content achievement between 50 to 74 percent.
The findings revealed that learners of EMI classes quite negatively reflect the content comprehension in EMI lectures. Due to the lack of academic knowledge achievement, learners show their preference in lectures taught in their mother tongues than in English (Al-Kahtany et al., 2016). Furthermore, students in countries that English is a foreign language were found to face more challenges in lesson comprehension in EMI classes that those who live in countries English is more pervasively used (Keuk & Tith, 2013). The lack of adequate English level may make students unwilling to take EMI classes because they perceived that taking EMI classes substantially hindered their knowledge absorption (Kim & Kim, 2014). Table 3 shows ten different factors hindering students' lesson comprehension in EMI classes using 5-point Likert-type scale from 1(Strongly disagree) -5 (Strongly agree) with 3 representing neither agree nor disagree. It is surprising that the mean score of 7 out of 10 difficulties listed are under theoretical midpoint (M = 3.0). The figures indicate that Hoa Sen University students face quite a few difficulties in EMI classes. The findings show that the variable -I receive less interaction and feedback from lecturers‖ (M = 2.55) is the lowest. This shows the difference with Joe and Lee findings (2012) when 74% Korean engineering students reported they received lack of lecturers' interaction and feedback. This could be the result of the university's long-term strategy setting qualified teaching as the highest priority and instructors are equipped with teaching pedagogical approaches through trainings, discussions, workshops, and symposiums. Some other findings are -I cannot listen and comprehend lectures‖ and -I cannot read and comprehend textbooks and related materials‖ with M = 2.75 and M = 2.82 respectively. According to Keuk and Tith (2013), university students often struggled with academic reading activities. Besides, Kagwesage (2013) found out 77% students admitted their listening skills were insufficient and 98% EMI instructors believed students were unable to listen English lecture comprehensively. However, the findings show Hoa Sen University students seemed to have fewer difficulties in reading and listening. Some others factors with higher mean scores are -I have difficulties to access related materials‖ (M = 2.86), -I cannot digest a great amount of content knowledge in limited time‖ (M = 2.87) and -I cannot enjoy the lesson because of passive teaching method‖ (M = 2.88). The results show a lack of agreement with previous researches when passive teaching methods displayed the most problematic for EMI students. In Kim, Kweon and Kim (2017)'s study, very few students thought their EMI classes were as interactive or satisfactory as their classes conducted in their native language. Nevertheless, the mean score of -I cannot understand lecture because of lecturers' poor pronunciation and fast speed‖ (M = 2.92) is noticeable. This was supported by study of Kim, Kweon and Kim (2017) that more than 40% Korean research students were positive with their instructors' language ability while only around 10% evaluated it negatively. The finding also indicated that lecturers' oral skills are not a significant problem inhibiting students' understanding their lessons. Kagwesage (2013) indicated that students' speaking abilities are not sufficient for EMI courses as they are only able to produce ‗single word utterance' (p.66). Moreover, students tend to be less confident with their speaking skills (Keuk & Tith, 2013). Similarly, the findings of this research also reveal that most students have difficulties in speaking full sentences with average speed and accurate grammar in EMI courses with M = 3.11. The highest difficulty that the research participants face in the study is understanding vocabulary in textbooks and lectures with mean score = 3.47. The findings show the similarities with Chang (2010) and Rogier (2012) which asserted students have problem with understanding EMI lessons because of insufficient vocabulary knowledge. Finally, -I cannot write a logical paragraph/ essay with various use of vocabulary and critical thinking‖ (M = 3.5) was found as the most difficult factor for students to deal with in EMI class. Likewise, Al-Kahtany et al. (2016) pointed out Arabic university students lacked sufficient writing skills in terms of ideas, critical thinking, and vocabulary. Standard multiple regression analysis was implemented to explore the correlation between variables of difficulties students faced in EMI courses and the overall EMI lesson comprehension and to evaluate how well items of difficulties could significantly predict the research participants' lesson comprehension. The results indicated that the model was statistically significant, F(10, 216) = 14.994, p <.001, and explained for 41 percent of the variation for the EMI lesson comprehension (R 2 = .410). The results from the regression indicated that, among the variables added, three of them did significantly predict value of lesson comprehension. These included -I cannot speak full sentences with average speed and accurate grammar‖ (Beta = -.303, t= -3.154, p < .05), -I cannot listen and comprehend lectures‖ (Beta = -.471, t= -3.516, p < .05), and -I cannot digest a great amount of content knowledge in limited time‖ (Beta = -.316, t= -2.487, p < .05). The findings indicated that the inability of listening and comprehending lectures makes the strongest contribution in explaining the dependent variable when controlling other variables in the model. This shows that listening skill plays the most important roles in hindering the learners in comprehending the lesson content in EMI classes. The result is in line with Kim, Kweon and Kim (2017) and  that listening skill presented the largest problems that EMI students' face and low English listening competence hampered EMI students from understanding their instructors' explanations, leading to failure of the content knowledge comprehension.

The Correlation between Variables of Difficulties Students Faced in EMI Courses and EMI Lesson Comprehension
Beyond the findings from quantitative data analysis, data from the qualitative data showed some additional worth-noting causes leading to the difficulties in comprehending EMI lessons. First and foremost, insufficient vocabulary range was mentioned by a couple of the research participants to be one of the most influential factors leading to the lack of lessonfully understanding. Research informants showed their dissatisfaction about the great amount of specialized words in EMI courses. This made them struggle with the lecture. One student revealed: "EMI courses make me overwhelmed as there are too many specialized words that I have to learn and remember. Even when I have not remembered the previous words yet, I have to get used to with new ones." However, responses among the informants were not consistent. While some believed that students should be taught common words useful for daily conversation, some claimed that thanks to EMI courses, they could learn many useful words and expand their vocabulary bank for understanding lessons effectively.
Inadequate writing skill was another noticeable issue. Students participating into the study admitted that insufficient writing skills affected their lesson comprehension. Particularly, note-taking considered as the biggest hurdle of writing skills hindered their comprehension. Some expressed they did not know how to note down key words and main ideas of lectures. Some interviewees raised their voice "Before taking EMI course, school should employ some trial classes to let students practice necessary study skills, so they could handle the content knowledge better". Meanwhile, other participants complained the way their lecturers presented the Power-point slide containing too many words made it hard for them to figure out the important points to note down "I suppose my lecturer should show the Power-point slide including only main content and illustration images".
The way lecturers delivered the lessons also caused difficulties for lesson comprehension. The participants claimed that the problems associated with lecturers' oral skills such as the fast speed when delivering lectures and the poor vocabulary usage hinder their comprehension during class time. One participant mentioned "I have tried to understand the lessons as much as I can but sometimes my lecturers talk too fast for me to catch up with the content." This is supported by Goodman (2014) and Kagwasage (2012) that rapid instruction speed would pause students' understanding. Another student supposed that professors should adjust their speed suitably based on the importance extent of the content "I think lecturers should slow down and repeat the significant ideas instead of surfing the content too fast". Besides, students hoped their lecturers could use a wide range of vocabularies in order to make the lessons clear rather than sticking to the content in textbook which is literally hard to obtain "I would like my lecturers use different words from the textbook to make the content sound familiar and easy for us." Teaching methods were also found to play an important role in comprehension of Hoa Sen University students. This finding is similar to the studies of Goodman (2014) and Hu and Lei (2014). Some interviewees, for example, were dissatisfied with the lecture-style and teacher-led teaching methods in their current EMI courses as these teaching approaches hindered their comprehension. These interviewees revealed that their lecturers basically explained the lesson content composed on the Power-point slides, extracted from the textbook without creating interactive or inquiry-based activities. Therefore, they tended to feel bored with the passive ways of receiving knowledge and made no effort at understanding the lecture. This is indicated by some informants as "I wish my class would contain many interesting activities so that whole class would be more excited and engage in the lesson" and "I would like to join in class activities because I think it is the best way to learn".

CONCLUSION AND IMPLICATIONS
The research aimed to find out the factors hindering students' lesson comprehension in EMI classes at Hoa Sen University. It also attempted to look for the extent that variables of difficulties could predict EMI lesson comprehension. Descriptive statistics revealed that inadequate writing skills, limited vocabulary range, and inadequate speaking skills are three major challenges to students. Particularly, multiple regression analysis pointed out that language skills, including listening and speaking, together with content knowledge are those of making contribution in explaining the understanding ability of the students in EMI classes.
This study has highlighted the factors of current practices of EMI implementation hindering the learners in lesson comprehension in general. However, the study failed to investigate each aspect profoundly and meticulously and make a comparison between the benefits and costs that the learners gain and pay between EMI classes and classes taught in their mother tongues. Future researches should consider researching deeply and extensively each aspect such as lecturers' perspectives, learners' characteristics, EMI approach policies and other potential aspects at different university contexts. Future researches would also make a great contribution if taking benefits and costs that the learners gain and pay in EMI classes and classes taught in their mother tongues into careful consideration.
This study can be a reliable source for the relevant stakeholders to consider how to design an appropriate EMI curriculum and take account of EMI policies that not only meet the students' needs but also the demand of EMI implementation at HEIs. Since students complained about the gap between their own English abilities and English level required in EMI courses -Sometimes I found the course is not for students who had low English proficiency, I prefer learning this course whenever my English is good enough‖, EMI courses of specific subjects should be organized in accordance with the students' language abilities, from low to high proficiency level in a professional manner.
The study has showed that it is necessary to provide academic assistance such as EMI course orientation, language skills center, EMI learning forum, etc. for students to better equip them with both the language and knowledge needed. Recommendations and expectation of participants are that "if I were prepared with a short course of how to learn EMI subjects, I believed I could deal well with any EMI courses‖ and -I would like to have a teaching assistant in my class, so whenever I could not keep up with the lesson I could ask again‖.
To limit the challenges that students face in EMI classes, instructors should be more flexible to adapt appropriate teaching methods for EMI curricula. The instructor should be highly aware of their students' troubles in EMI courses in order to develop alternative efficient teaching techniques to obtain students' lesson comprehension and inspire students with language learning. One of students' shares -The passive teaching method makes me bored and exhausted while attending the EMI class. I wish my lecturer could create some activities to warm the environment up and make students eager in lessons‖ and -I think teaching styles influences too much how students understand and enjoy the lessons‖.