Language Teachers’ Perception and Practice of Adopting Blended Learning to Adapt to the New Normal

The practice of blending online learning and face-to-face learning has become ubiquitous across many tertiary institutions both worldwide and nationwide, especially amid the Covid-19 pandemic. This paper is based on the part of a study investigating how language teachers perceive blended learning and how they adopt this model in their teaching practice to adapt to the new normal. The participants include 50 teachers who are teaching languages such as English, French, Russian, Chinese, Japanese, and Korean at Hue University, located in central Vietnam. The data was collected through a survey and then supported by follow-up interviews with 10 of the surveyees. The findings reveal teachers' perceptions of blended learning in terms of its necessity, feasibility, usefulness, and ease of use, as well as their confidence and intention of continuation. In addition, the study also reflects the reality of how blended learning is currently applied by these language teachers regarding such factors as how the online component supports the face-to-face component, how online and face-to-face classroom activities are blended, when online and face-to-face elements are arranged, etc. There is also a comparison of these teachers’ responses based on their teaching expertise to see whether they perceive and use blended learning differently.


Introduction
The advent of information communication technology (ICT) has exerted a plethora of significant impacts on every aspect of social life, the educational field included. The advancement of educational technology has shifted the teaching paradigm and opened more opportunities for both teachers and students. As Hofmann (2018) commented, the modern classroom is not necessarily just a place, and learning should be an experience that can flexibly take place anywhere and at any time. The idea of organizing a virtual learning environment in addition to the conventional physical classroom, also known as blended learning, has emerged as one of the greatest educational changes in the last decades. In fact, the practice of blending online learning and face-to-face learning has become ubiquitous across many tertiary institutions both worldwide and nationwide, especially amid the Covid-19 pandemic when changing and adapting is the only way for educators to maintain education (Bordoloi et a., 2021;Nguyen & Nguyen, 2021;Tran & Nguyen, 2022). In language education, blended learning has been implemented for a long time as it is believed that language teachers and learners are in true need of a learning environment which is wider and more flexible than the traditional physical classroom so that language acquisition and practice can be better enhanced (Hubackova et al., 2011;Dennis, 2013).
At the University of Foreign Languages and International Studies, Hue University, which is a regional university located in central Vietnam, before the global pandemic, a humble number of teachers at one or two faculties were familiar with the blended learning concept and applied this model in their teaching practice (Hoang, 2015;Cao, 2017;Phan, 2018). However, since 2020, as an immediate response to the drastic effects of the Covid-19, all teachers across every faculty of the institution were firstly encouraged and then urged to implement online learning besides traditional face-to-face learning in order to support students. It is widely accepted that what teachers believe or think is important because their beliefs and attitudes toward innovation can impact their application in reality (Gilakjani & Branch, 2012). Therefore, this paper, which is part of a larger study, aims to investigate the teachers' perceptions and practices of implementing blended learning in language education from 2020 until now.

Definition of blended learning
There are a variety of approaches to defining what blended learning is as it has become "an umbrella term" (Hrastinski, 2019, p. 564). Literally speaking, "blended" indicates the process of mixing or combining two or more components together to achieve a desirable mixture. In the "learning" context, blended components are often instructional activities or events that aim at increasing teaching and learning effectiveness.
From a broad perspective, Driscoll (2002) mentioned four dimensions of blended learning, including 1) combining modes of web-based technology, 2) combining various pedagogical approaches, 3) combining any forms of instructional technology with face-to-face instructorled training, 4) combining instructional technology with actual job tasks. Blended learning can also be approached from different perspectives, such as a holistic perspective (integrating instructional media into a traditional classroom), an educational perspective (combining online with traditional face-to-face class activities), or a pragmatic perspective (mixing different pedagogical strategies) (Kaur, 2013). Likewise, Hofmann (2018) claimed that blended learning is about "aligning learning objects with the most appropriate instructional strategies, techniques, and technologies while meeting the need of the organization and modern learners" (p.15) However, blended learning is often most popularly known for its narrower meaning. Graham (2012) defined that "blended learning systems combine face-to-face instruction with computermediated instruction" (p.5). With this definition, he emphasized four factors of the face-to-face and distributed learning environment, which can be mixed, including space (physical versus virtual), time (synchronous versus asynchronous), fidelity (high versus low), and humanness (high human, no machine versus no human, high machine). Staker and Horn (2012) more specifically defined blended learning as follows: Blended learning is a formal education program in which a student learns at least in part through online delivery of content and instruction with some element of student control over time, place, path, and/or pace and at least in part at a supervised brickand-mortar location away from home. (Staker & Horn, 2012, p. 3) The definition of blended learning proposed by Graham and Staker and Horn is in line with other definitions held by other researchers like Hockly (2018), Hofmann (2018), Anthony et al. (2019), who agreed that blended learning is the combination of the face-to-face learning ACOJ-ISSN: 1936-9859 AsiaCALL Online Journal Vol. 13;No. 5;2022 31 environment with a virtual learning environment. In order to achieve effective outcomes, the combination between face-to-face learning and online learning should be done in a systematic (Bliuc et al., 2007), thoughtful (Alammary et al., 2014), planned and pedagogically valuable manner (Picciano, 2009). This study also adopts the approach that blended learning indicates the hybrid instructional delivery in which on-site learning activities are combined with online learning activities in order to provide students with more learning opportunities beyond the brick-and-mortar classroom.
The combination of face-to-face learning and online learning in a blended model, which (Bleed, 2001) referred to as a 'bricks and clicks' model, can involve a number of components. Kaur (2013) mentioned three main components of blended learning: learning environment, which can be either synchronous or synchronous, media which refers to vehicles used to deliver learning content, and instruction which means the selection of appropriate teaching strategies to support the learning objectives. Meanwhile, Alammary (2019) listed five basic components which constitute blended learning based on the types of interaction between teachers and students in both offline and online channels. The five components are face-to-face instructorled, online instructor-led, face-to-face collaborative work, online collaborative work, and online self-paced. Depending on how these components are blended to achieve teaching and learning goals, there are different types of blended learning which will be discussed below.

Types of blended learning
There are a number of types of blended learning depending on the different categorizing criteria. This study deploys the approach of Staker and Horn (2012), and Graham (2012) to understand different types of blended learning that teachers have applied in their teaching context in terms of how and why face-to-face learning and online learning are blended.
Based on the schedule of face-to-face and online components, Staker and Horn (2012) divided blended learning into four types. The first type is Rotation Model, in which, within a given course teaching, activities are rotated on a fixed schedule or at the teacher's discretion between face-to-face and online modalities. The rotation Model is also divided into some sub-categories, including Station Rotation which means students rotate on a fixed schedule among classroombased activities (or stations), at least one of which is done online; Lab Rotation which involves students rotating on a fixed schedule among classrooms and computer labs on the campus; Flipped classroom in which students rotate on a fixed schedule between teacher-guided practice in brick-and-mortar classrooms and online-delivered instruction; and Individual Rotation in which students rotate on an individually customized fixed schedule among face-to-face and online modalities. The second type is Flex Model; with this model, learning content is mainly delivered online, and students can study on an individually customized flexible schedule with face-to-face support when needed. The third type is the Self-Blend Model involving students deciding to take extra online courses to supplement their face-to-face courses based on their individual needs. In the final type, which is called Enriched-Virtual Model, students divide their time to attend face-to-face classrooms and online-delivered content within the same course. Among these four models, Rotation Model and Enriched-Virtual Model seem to fit best into the context where the study was conducted because all the undergraduate courses are campusbased, online activities can be added to the traditional curriculum, but the learning and teaching have to be based on a fixed and planned schedule.
Graham (2012) examined the primary purpose of blended learning and divided blended learning into three types. The first type is Enabling Blends which aims at providing students with additional learning activities of the same learning experience but through a different modality. According to Alammary et al. (2014), this way of blending two learning modalities can have a low impact because teachers do not have to make large changes to the traditional curriculum. The second type is Enhancing Blends bringing some incremental changes to existing pedagogy by offering online materials to replace some face-to-face activities without significantly changing the way teaching and learning take place. The replacement of traditional learning activities in face-to-face classrooms with online activities can have a medium impact on the existing curriculum (Alammary et al., 2014). The third type of blended learning, which has a high impact, is called Transforming Blends involving significant changes in pedagogy through incorporating web-based technology and redesigning classroom activities so that online and offline learning can support each other.
Technology Acceptance Model and blended learning Gilakjani and Branch (2012) once said that "Teachers' beliefs are essential in considering how a teacher teaches, thinks, and learns" (p.62) and that if teachers have positive attitudes toward technology integration, they can enhance the use of technology in teaching and learning; otherwise, their negative attitudes can constrain it. Kannia (2014) added that the success of implementing technology in education significantly depends on the teachers' acceptance level, so their readiness should be seriously taken into account. In order to understand the teachers' attitudes toward and their integration of technology in classrooms, Technology Acceptance Model (TAM) proposed by Davis (1989) has been the most widely used (Teo et al., 2018).
Originally, TAM proposed that the users' perception of ease of use and usefulness of technology are predictors of their attitude towards using the technology, subsequent behavioral intentions, and actual usage. Specifically, perceived usefulness refers to the extent to which the users believe that using the technology will improve their work performance, while perceived ease of use refers to how easy they think using the technology will be. Usefulness and ease of use are considered two important factors which can impact the users' attitude towards using the technology, and such attitude towards using the technology will determine their behavioral intention to use it in their practice. In the blended learning scenario, TAM implies that if teachers perceive that applying blended learning is useful for their teaching job and that it is easy for them to implement blended learning in reality, these positive attitudes can lead them to the actual implementation of blended learning and can determine how well they do that.
Some scholars later extended TAM by suggesting the addition of some other external factors which can determine people's acceptance and use of technology more effectively. Venkatesh et al. (2003) suggested adding facilitating condition, which is defined as "the degree to which an individual believes that an organizational and technical infrastructure exists to support the use of the system" (p. 453). In the context of blended learning, Facilitating Condition means teachers' perception of whether they have the necessary resources to apply blended learning, such as technological facilities. Abdullah and Ward (2016) also revealed some other factors, such as self-efficacy and enjoyment of the original TAM. Self-efficacy refers to the teachers' judgment of their own capability to implement blended learning. Meanwhile, the concept of enjoyment is the extent to which blended learning is perceived to be enjoyable or interesting to try. This study also adopts TAM as the core theory, which will guide the research to investigate teachers' perceptions and practices of blended learning. In order to understand teachers' perception of blended learning, this study refers to these factors of TAM and Extended TAM in designing research instruments.
A number of studies were conducted to investigate language teachers' perceptions and practices of blended learning (Rivera, 2019;Ismayana et al., 2020;Saeed, 2020). In this study, more attention is paid to the studies in which TAM was applied in order to understand teachers' perceptions as well as their practice of blended learning at the tertiary level. Rahman et al. ACOJ-ISSN: 1936-9859 AsiaCALL Online Journal Vol. 13;No. 5;2022 33 (2019) conducted research adopting TAM to find out the attitudes of 206 ESL Malaysian teachers toward the use of flipped learning, one of the blended learning models. This quantitative study focused on investigating the relationship between computer self-efficacy and computer anxiety and lecturers' attitudes toward the flipped classroom. The findings showed that there was no significant relationship between teachers' computer competence and their attitudes toward flipped learning, which can help to understand their acceptance behavior of this model. Sánchez-gómez et al. (2020) conducted a study to collect data from 982 Spanish lecturers in order to analyze their beliefs, expectations, and attitudes regarding the acceptance -and adoption -of blended learning from both quantitative and qualitative perspectives. The results showed that the variable intention essentially determines the lecturers' acceptance of the use of blended learning.
These aforementioned studies show that there is a relationship between TAM variables and teachers' perceptions of blended learning and that teachers' perceptions can impact their acceptance and use of blending learning as an innovative teaching methodology. However, little is known about their practice of adopting blended learning, which subsequently took place after they perceived the use of blended learning. Therefore, this study was conducted with the purpose of finding out more information about teachers' perceptions of blended learning based on variables of TAM and their practice of adopting blended learning as a teaching approach. In addition, its ambition was also to see whether there is any relationship between teachers' teaching expertise and their perceptions and practices of blended learning or not.
With these purposes in mind, the study would like to seek the answers to the following questions:

Research questions
1. What is language teachers' perception of blended learning? 2. Is there any difference in teachers' perception of blended learning regarding their teaching expertise? 3. How do language teachers apply blended learning in their teaching practice? 4. Is there any difference in teachers' practice of blended learning regarding their teaching expertise?

Research design
This study employs a mixed research method as the main approach to finding the answers to the above research questions. According to Dornyie (2007), the combination of quantitative and qualitative approaches can help yield "an enriched understanding by illustrating, clarifying, or elaborating on certain aspects" (p.164). The research design is QUAN => qual (Dornyie, 2007), in which a questionnaire survey is followed by an interview. Qualitative data collected through the follow-up interviews can add flesh to the bones by providing more in-depth explanations for the survey respondents.

Research sample
Purposive sampling, which is defined as "a form of non-probability sampling in which decisions concerning the individuals to be included in the sample is taken by the researcher, based upon a variety of criteria" (Rai & Thapa, 2015, p. 5), is used to select the research sample. A filtering question was used to select the potential participants for this study. Accordingly, a number of teachers from different faculties of the institution were asked whether they had https://asiacall-acoj.org Cao Thi Xuan Lien Vol. 13;No. 5;2022 34 applied blended learning in their teaching context or not. Only teachers who knew about this concept and had real experience with it were invited to participate in the study. Quota sampling (Dornyie, 2007) is also employed to decide the number of teachers participating in the study was also selected based on the scope of each faculty. Fifty teachers who met the preliminary condition participated in the survey, and then 10 of them agreed to join the follow-up interviews to share more about their perceptions and practice of blended learning. The demographic information of these participants is summarized in the table below. As can be seen from Table 1, the majority of participants (taking up 84%) are female, and over 50% of them have been teaching different languages for at least 10 years. The Faculty of English and the Faculty of Chinese had a larger number of teachers participating in the survey as these two faculties are the biggest faculties at this institution.

Research instrument and procedure
The data was collected through an online survey using Google Forms and online interviews using Zoom. The reason for carrying out the data collection online was that it was a convenient way to reach a large number of teachers who had different busy teaching schedules. The questionnaire consisted of 10 Likert-scale items, which were purposefully designed based on a number of TAM variables (Davis, 1989;Venkatesh et al., 2003;Abdullah & Ward, 2016) to collect information about how teachers perceived blended learning. In addition, four multiplechoice items, which were designed according to Staker and Horn's (2012) and Graham's (2012) classification of blended learning types, aim at finding out how they implemented blended learning in practice.
The online survey was first piloted with three senior teachers and then revised according to their feedback to ensure its clarity and consistency. The data collection procedure lasted for two weeks in late July 2022, when the survey link was shared with the participants via email and social media platforms after they had been informed about the purpose of the study. Following the survey is a structured interview conducted via Zoom to get in-depth responses from some of the participants to clearly understand their perceptions and practice of applying blended ACOJ-ISSN: 1936-9859 AsiaCALL Online Journal Vol. 13;No. 5;2022 35 learning. After completing the survey, the data were analyzed using SPSS and then presented in tables and charts.

Language teachers' perceptions of blended learning
In the survey, the participants were asked to share what they thought about blended learning by choosing one of the five options, including Strongly to disagree (=1), Disagree (=2), I have no idea (=3), Agree (=4) and Strongly agree (=5) for each of the given statements written based on different TAM variables. As listed in Table 2 below, there were ten Likert-scale statements in the survey which had been created based on such extended TAM variables as perceived usefulness (1-for teachers and 2 -for students), perceived ease of use, enjoyment, facilitating condition, self-efficacy (in terms of teaching and technology), and intention of continuance. In addition, the research also added necessity and feasibility as two other variables which can help to clarify teachers' perceptions of blended learning in the study context. Teaching self-efficacy 10 I will continue applying blended learning in the future.

Intention of continuance
The calculation of their responses in terms of Mean and Standard Deviation using SPSS 29 is summarized in the following table and chart.

Figure 1. Survey results about teachers' perception of blended learning based on different variables
It can be seen from Table 3 and Figure 1 above that the Mean values range from 3.08 to 4.20, which indicates that all of the teachers in the survey tended to agree and strongly agree with the statements about different aspects of blended learning. This can reveal the fact that they had a positive attitude toward blended learning. The comparison of the teachers' responses to the surveyed variables, as shown in Figure 1, reveals that such aspects as the necessity, feasibility, and usefulness of blended learning received a higher appreciation (Mean > 4.0) than the others (Mean < 4.0). Meanwhile, ease of use was the variable that had the lowest consensus (Mean = 3.08). This finding is quite surprising because it can be seen that despite the fact that teachers' self-efficacy in terms of both technological and pedagogical knowledge and skills, as well as their perceived facilitating condition in terms of technological facilities both at home and at school, were high (4.0< Mean >3.5), they still had a reluctant attitude to the idea that it was easy to apply blended learning.
The responses of some teachers in the follow-up interviews about why they thought applying blended learning was not easy to have provided some insights into this issue. Below are some extracts from the teachers' answers in the interviews: Such difficulties as students' lack of facilities, increase in time consumption and workload, and large classes were also the common themes that many teachers mentioned in their answers to explain why they perceived blended learning was not easy to implement. The findings from both survey and interview data reveal that although teachers in the study held a positive attitude toward blended learning, and they especially appreciated some aspects like its necessity and feasibility to be implemented at their institution, its usefulness for not only their teaching but also students' learning and its enjoyment. Additionally, teachers also perceived that they had enough technology, teaching self-efficacy, and facilitating conditions to help them apply blended learning in their teaching context. However, some other aforementioned external factors also caused difficulties which lowered their perception of blended learning ease of use.
In order to find out whether there is any difference in teachers' perceptions of blended learning in regard to their teaching expertise or not, a correlation analysis is also run to test the relationship between teachers' years of teaching and language of teaching with the Mean value of their responses to the questionnaire items above.
https://asiacall-acoj.org Cao Thi Xuan Lien Vol. 13;No. 5;2022 38 -tailed) .329 N 50 50 As indicated in the above table, there is almost no difference in teachers' perceptions of blended learning in regard to their years of teaching, with r = 0.02. In addition, in terms of their language of teaching, there seems to be an insignificant difference with r = 0.141 in the way teachers from different language faculties perceived this teaching method.

Teachers' practice of implementing blended learning
To investigate the teachers' practice of implementing blended learning in their teaching context, the study aims at collecting information about the following aspects: (1) when face-to-face learning and online learning are blended, (2) what the purposes of blending face-to-face learning and online learning are, (3) how teaching activities are selected to be organized in faceto-face and online modalities.
The survey findings are summarized in the following bar charts and followed by the interview responses: As can be seen from the chart, the most popular ways to blend face-to-face and online activities that teachers from most faculties chose is that students attend traditional classes on prescribed timetables, and then they were asked to participate in some other online activities at their convenience either after or before the class time. In addition, a considerable proportion of teachers also decided to arrange face-to-face lessons and online lessons at a fixed time. Meanwhile, the other two ways of blended learning were far less favorable.
It can be noticed that the three most prevalent types of blended learning that most teachers have applied have one thing in common; that is, all the learning activities strictly followed a preplanned timetable. To justify this tendency, some teachers shared in the interviews that they prioritized the consistency of learning time because "it is convenient for both teachers and students to work according to the timetable" (T16 from Faculty of Chinese), "it can make sure teachers and students complete the teaching and learning plan regulated by the university" (T10 from Faculty of English), and "it helps students manage their learning actively" (T19 from Faculty of International Studies). Moreover, some teachers also explained why they chose to add online activities to the classwork. A teacher from the Faculty of ESP (T11) said, "the theoretical content often consumes much class time, so providing students with online lessons in advance can help students have more time in class to practice. Language skills like writing will be more effective if students can practice with the teacher's guidance".
The chart below shows the teachers' responses about the purpose of adding online activities to the traditional curriculum: As revealed in the chart, most of the teachers indicated that the most popular type of blended learning that they applied is Enabling blend because their main purpose of using online activities was to supplement what students had learned in class. Offering extra activities on the Internet in order to help students consolidate in-class lessons was also the main theme that many teachers mentioned in their interview answers.
"Helping students to consolidate the knowledge they have learned through computerbased exercises can be beneficial because students will feel more interested and can access the online materials anywhere and anytime".  Cao Thi Xuan Lien Vol. 13;No. 5;2022 40 "Providing extra online activities will be suitable for some students who cannot keep pace with the in-class lectures."

(T11 from Faculty of ESP)
"Teachers can impart students with foundational knowledge in the traditional classroom, and then give them more exercises online to help them revise the knowledge".
(T32 from Faculty of Japanese) " The time in class for each lesson is often quite limited (only two 45-minute periods/week), so providing students with supplementary knowledge on the online platform can effectively support their learning".
(T10 from Faculty of English) More than 50% of teachers decided from the beginning of the course or the semester about which learning contents would be taught in class and which ones would be delivered online. In so doing, online lessons played an as equal role as face-to-face lessons in delivering teaching content to students throughout the semester. The following are some comments made by the teachers about this way of blending two learning modes: "Some teaching activities are only effective when they are organized in face-to-face classrooms; meanwhile, some others are more interesting and successful if they are taught online." (T46 from Faculty of English) "Identifying the teaching plan from the beginning helps both teacher and students to make a clear plan and preparation for the whole semester." (T10 from Faculty of English) As shown in the chart, 40% of teachers chose to replace some face-to-face learning activities with online activities. A teacher from the Faculty of Chinese said that: "I replace some face-to-face teaching contents with online activities because I find it easier for some activities such as games and web-based quizzes to be organized online." (T16 from Faculty of Chinese) The following chart presents teachers' priority for organizing learning activities in face-toface and online learning environments. It is demonstrated in the chart that some learning activities, such as discussion, lectures, and group presentations which require a high level of teacher-student interaction, were more favorably organized in the face-to-face classroom than online. In addition, more teachers prioritized conducting high-stake summative assessments such as mid-term tests in conventional classrooms where they could have more control over students' activities. In contrast, for low-stakes formative assessments through exercises or assignments, more teachers would like to organize them online because of their convenience and immediate feedback. Online platforms were also preferred by most teachers in the study as a place where they could share a variety of learning materials with students.

Discussion
The findings reveal that, generally, teachers in the study had a positive perception of blended learning, which is in line with the findings of previous studies conducted by (Rivera, 2019;Ismayana et al., 2020;Saeed, 2020). Specifically, the teachers indicated a high level of agreement with the necessity and feasibility of implementing blended learning in their institution. Most teachers also highly appreciated this teaching method's usefulness for their teaching job and students' learning process. Like the results found by Rahman et al. (2019), teachers in the survey showed that their perception of their own technology self-efficacy was high, which means that there could be a good prediction of them applying blended learning in their practice. However, despite the fact that teachers had a high level of self-efficacy and facilitating conditions that can support their implementation of blended learning, it seems that they still found it uneasy about applying it in reality because many of their students lacked needed facilities, and this teaching mode increased their workload. As shown by Sánchezgómez et al. (2020), teachers also admitted that blended learning requires a great amount of time and effort, which can be a possible hindrance. It can be said that their positive attitude towards blended learning can help to predict their actual use of blended learning in practice (Davis, 1989). However, taking into consideration some unavoidable obstacles related to Face-to-face https://asiacall-acoj.org Cao Thi Xuan Lien Vol. 13;No. 5;2022 42 students and their working context, these teachers might cautiously apply blended learning. Moreover, the comparison of teachers' perceptions across faculties based on their teaching expertise also shows an insignificant difference. This can be said that at the study time and in the study context, there is no clear relationship between what they teach and how they perceive teaching it in a blended learning approach.
In terms of teachers' practice of blended learning, the above findings suggest that, compared with the four blended models suggested by Stake and Horn (2012), it seems that at the institution where this study was conducted, some blended learning models are more in favor than others. It can be seen from the finding mentioned above that Rotation Model and Enrichedvirtual Model are the two popular types of blended learning that most teachers apply in their practice. Sadeed (2020) also found that teachers often prefer the Rotation Model of blended learning, especially station rotation and flipped classroom models, because it is easy to implement as a beginner or initial teacher who is trying to incorporate technology in the classroom. These models are similar in that they are based on a fixed schedule, and face-to-face instruction remains crucial in the learning process, while online instruction and activities play a scaffolding role. Meanwhile, concerning the purpose of blending two learning modes, in reference to the models which Graham (2012) proposed, it can be said that most teachers applied 'enabling blend' in their practice. In this blended learning model, online activities' main role is to supplement students with extra learning materials and activities to help them better acquire what they have learned in the traditional face-to-face classroom. According to Almamary et al. (2014), the approach in which the online component is used to supplement face-to-face lessons is the simplest way of blending online activities into the traditional curriculum and has a low level of impact on the existing teaching and learning process. Enabling blends also requires teachers to develop less complicated technological knowledge and skills, so they can easily apply them. Additionally, it is also indicated that teachers had a tendency to prioritize face-to-face class time for important teaching activities such as lectures, collaborative work, and formal assessment. Meanwhile, online platforms were preferred for other activities which needed more flexibility and convenience. These findings about teachers' blended learning practice align with their positive but cautious perceptions, as mentioned above.

Conclusion
As blended learning can maximize the advantages of both face-to-face and online learning, it will definitely become the tendency of future education. Therefore, investigating how teachers who are the indispensable stakeholders in this paradigm shift perceive and implement blended learning, in reality, is of necessity. This study has shown that teachers' positive perceptions of the benefits of blended learning can lead them to active application in practice. However, because teachers still held a positive but worrying attitude towards blended learning in regard to some challenges that their students may have or the work burden that they have to shoulder when implementing this teaching method, they tended to be more reluctant and cautious in applying blended learning in their teaching practice. Therefore, in the current study context where traditional classroom-based curriculum still plays a primary role, students' access to facilities is limited, and teachers have to deal with large classes and heavy workloads, the addition of online activities with the main purpose of supporting classroom activities and enable students to learn beyond the brick-and-mortar classes is acceptable and worth encouraging. In the future, in order to motivate teachers to bravely implement blended learning at a higher level of complexity and impact, the university must provide sufficient guidance, facilitate conditions as well as encourages policies for both teachers and students.

Limitations and future research
Although this study was conducted on a whole-university scale, the number of teachers (N = 50) participating in the survey and follow-up interviews remained humble in comparison with the total number of the entire academic staff (N > 250). Therefore, the findings may not accurately generalize the current context of how blended learning is perceived and applied at the institution. In addition, the study also adopted a certain number of variables in the Technology Acceptance Model and its extended version to determine teachers' perceptions of blended learning. More variables such as computer anxiety, experience, subjective norm (Abdullah & Ward, 2016), costs, demographics, and social influence (Venkatesh et al., 2003) should have been added to the questionnaire items so that a more in-depth understanding of teachers' perceptions could be attained. This study is also limited in that it only compares teachers' perceptions of blended learning in terms of their teaching expertise, so further research can take into consideration some other aspects, such as teachers' genders, educational background, technology experience, etc., to see whether there is any relationship between these factors and how teachers perceive and practice blended learning.