Introduction

Two years with the Covid-19 pandemic, the most worrying thing is the disruption in education, leading to the education industry has to change many training and teaching methods, which online learning will be more sustainable for the new situation (Adedoyin and Soykan, 2020). According to UNESCO data, 290 million pupils have been obstructed in their studying due to the school and campus closures caused by the Covid-19 virus (Purwanto et al., 2020). This move threw practically everyone’s life into disarray, especially those working in educational institutions (Bamoallem and Altarteer, 2021). Thus, Martinez (2020) argued that there has been an increasing trend toward digital learning. From the face-to-face classrooms to classrooms on the online application, from the interpersonal to the virtual learning, and from the offline conferences to the webinars, online methods have become a pedagogical movement from old methods to new approaches to teaching and learning (Mishra et al., 2020). Blended learning was applied in the incipience of Covid-19 and after overcoming the peak of Covid-19. By maintaining the present hygienic regulations of social distance and seating capacity, blended learning at this time enabled a partial restoration to the intended “normality” (Batista-Toledo and Gavilan, 2022). Because this modality allowed for the inclusion of engaging information in face-to-face sessions and productive contact with the instructor, students felt that blended learning provided a learning environment that was satisfactory (Peimani and Kamalipour, 2022).

Despite the fact that most public institutions and some private universities in Vietnam currently provide online education and distant learning deanships during the Covid-19 pandemic. While the majority of faculty members have no experience in online teaching. Students and faculty were compelled to quickly adapt and change existing teaching practices in order to provide the entire course material online (Bamoallem and Altarteer, 2021). According to its perceived efficacy in delivering flexible, timely, and ongoing learning (Rasheed et al., 2020), blended learning has been regarded as the most effective and most popular method of teaching in the Covid-19 epidemic. Interestingly, the students emphasized the independence that blended learning gave them to schedule their time and pursue their studies at the rate that worked best for them (Buck and Tyrrell, 2022). This enhanced student freedom and engagement in their studies, which helped them do better academically (Peimani and Kamalipour, 2022). Therefore, the concept of blending educational materials with online interventions has shown to be an improvement amongst offline learning and online learning methods (Rasheed et al., 2020). When comparing blended learning versus virtual learning, the overall course satisfaction was better in the former (Finlay et al., 2022). It also expands the learning opportunities and creates a good educational environment for students (Jost et al., 2021; Wai and Seng, 2014), and enables them to explore a perceptual change (Bazelais and Doleck, 2018). Many studies have functioned in the past to better interpret the efficacy of mixed studying methods in many aspects throughout the world. In the Saudi region, the role of inquiry’s community framework was analyzed by Bamoallem and Altarteer (2021) to determine the predictors and whether there was a significant association between distant teaching experience as well as the perception of students and their acknowledgment of the blended learning method in the foreseeable as a stage in the progressive returning to daily life during the Covid-19 outbreak. They argued students’ perceptions and acceptance of blended learning were predicted by the instructional, cognitive, and social presence categories. While, in Malaysia, Rasheed et al. (2020) investigated the issues of blended learning’s online component from the perspectives of students, teachers, and educational institutions. As a result, they found the key issues that educational institutions were the provision of appropriate instructional technology and adequate teacher training support.

However, not many previous studies have explored students’ perceptions of the blended learning method and their behavioral intentions throughout the Covid-19 period in Vietnam. Besides, Vietnamese students have never experienced blended learning in the global pandemic when all schools encouraged their students to study online and opened a few traditional classes for some special subjects, which means blended learning was highly recommended during this period. From these arguements, it is necessary to set the research’ goal, which was to investigate the impact of blended teaching modules during Covid-19 on university students’ perspectives and behavioral intentions in the Vietnamese educational environment. Furthermore, the following research was undertaken through the moderating impact of the “Covid-19 understanding” variable, studies have been targeted at the implicit correlation between students’ perceptions and behavioral intention toward blended learning, which is the new research aspect. Besides, this study resulted in the conclusion that there is no relation between social presence and student perception and behavioral intention as well in the Vietnamese blended learning scenario, while the previous studies supported this finding. The reason for this new exploration should be considered in some aspects but the large contribution is the difference in the research area and the Covid-19 stage. Because this study targets Vietnamese students (in Ho Chi Minh City only) in various levels of education who have studied blended learning in the early stage of Covid-19 pandemic.

However, while many studies have explored blended learning’s efficacy globally, particularly in regions like Saudi Arabia (Bamoallem and Altarteer, 2021) and Malaysia (Rasheed et al., 2020), few have focused on students’ perceptions of blended learning in Vietnam during the Covid-19 pandemic. Given that Vietnamese students had not previously experienced blended learning on such a large scale, this study aims to fill that gap by investigating how blended teaching modules during the pandemic have impacted students’ perceptions and behavioral intentions in the Vietnamese context. Additionally, this study will explore the moderating role of students’ “Understanding of Covid-19” on their experiences with blended learning, a relatively unexplored variable in this context. Furthermore, the findings indicate that there is no relation between social presence and student perception and behavioral intention in Vietnam’s blended learning environment, a result that differs from previous studies. The variation may be due to differences in the research area and the timing during the pandemic.

The goal of this research is to contribute both theoretically and practically. Theoretically, the goal of this study is to broaden our understanding of the elements that influence students’ views and behavioral intentions toward blended learning, while also introducing the moderating role of the “Understanding of Covid-19” measure. In practice, the study intends to give educational institutions with insights into how to improve blended learning practices, increase student engagement, and limit any negative consequences such as lower enrollment and unhappiness. By evaluating Vietnamese students in Ho Chi Minh City who experienced blended learning during the early stages of Covid-19, this study offers a unique viewpoint on how blended learning might be enhanced in similar educational situations.

Literature review and hypothesis

The impact of distanced education during the Covid-19 outbreak on students’ perception of blended learning as well as behavior intention during the epidemic was very essential to research (Bamoallem, and Altarteer, 2021). Some researches were conducted in order to assess the variables that were related to influencing students’ blended learning experiences (Gopal et al., 2021; Tran, 2022). Thus, this study’s approach to the community of inquiry (CoI) framework serves as the theoretical foundation.

The community of inquiry (CoI) framework

According to Garrison, Anderson, and Archer (2000), the CoI framework was an online learning process paradigm. Swan (2019) showed that the CoI theory was a method model of e-learning and a blended environment in which the social construction of knowledge was rendered more difficult due to the time and space separation of course participants. Furthermore, Garrison (2015) also argued that the objective of the CoI framework was to investigate the effects of a purposeful education system on an individual’s perceptual processes. Besides, to create a meaningful educational experience, the CoI approach was used by Garrison et al. (2000), which emphasized three main aspects: cognitive presence, social presence, and pedagogical presence. The Covid-19 outbreak had a huge impact on educational practices, and online education within the CoI framework emerged as the primary solution (Martin et al., 2022). The CoI framework can assist to create a successful blended learning process by supporting with course design and the implementation of activities and tasks that develop students’ disciplinary knowledge and skills, stimulate their enthusiasm to learn, provide them with a greater sense of satisfaction with their progress, and encourage them to apply what they have learned to their professional studies.

Social presence

According to Lowenthal and Lowenthal (2010), social presence is a notion that describes individuals’ ability to represent themselves as actual people via a communication medium. Based on Short, Williams, and Christie (1976), these authors coined the term “social presence” to describe the various capabilities of multi-media for conveying nonverbal and audible signals, with the goal of expressing the affective and emotional (social) aspects of learning encounters in judgment. Gunawardena and Zittle (1997) defined social presence in mediated communication as the degree to which an individual is viewed as “actual”. Additionally, social presence was seen as having a direct influence on the formation of community and cooperation in online courses, which was an essential component of the CoI framework (Swan et al., 2009). Bamoallem and Altarteer (2021) demonstrated that the three sub-dimensions of social presence were group cohesiveness, open communication, and emotive reactions. Additionally, social presence acts as a partial bridge between instruction and cognitive presence and has a direct impact on the development of cognitive presence (Redstone et al., 2018).

Cognitive presence

According to Garrison et al. (1999), the level to which learners can develop and validate knowledge/meaning via module activities, argument, and assessment was associated with cognitive presence. Furthermore, Garrison et al. (2000) constructed the Practical Inquiry Model (PIM), which demonstrated a cognitive presence that encompasses the ability to generate and validate significance through continuous reflection. The PIM was separated into four phases: (1) In order to motivate students, the instructor highlighted the fundamental principle or subject in the form of a learning activity, which might include a challenge or dilemma; (2) Through planning meetings or individual research, investigation comprised getting a deeper understanding of the topic and seeking appropriate information and answers; (3) As students engaged in an analytical stage to consolidate their studying, the instructor explored for misconceptions. Almasi and Zhu (2020) identified an association between cognitive presence in a community of inquiry and perceived and real intellectual outcomes. Additionally, the authors found that almost all of the students in both courses expressed a strong sense of the course’s cognitive presence.

Teaching presence

Teaching presence was defined as the development, motivation, and direction of psychological and behavioral procedures with the goal of producing profoundly meaningful and intellectually appropriate academic accomplishment (Garrison, 2016). It comprises three components: (1) developing and structuring instructional materials, (2) fostering discourse, and (3) direct instruction (Anderson et al., 2001). Designing and organization refers to how teachers create learning activities, schedules, and other instructional elements for blended synchronous courses; facilitation discourse refers to how instructors facilitate and promote meaningful collaboration and discussion in mixed synchronous courses; Teachers must employ direct teaching to determine students’ needs and provide timely instruction in order to ensure that students achieve the necessary learning outcomes. (Fiock, 2020).

Blended learning

According to Ghazali et al. (2018), blended learning was first developed by various colleges in the United States and Canada in the late 1990s as a unique way of studying that combined face-to-face and online learning. It was characterized by Khalil et al. (2018) as a combination of traditional face-to-face instruction with virtual learning web-based. Driscoll (2002) expanded on integrated learning by mentioning four concepts: (1) blending internet technological modes (e.g., live stream access lessons, cooperative education methods, streaming video, audio, and text) to achieve the educational objectives; (2) educational techniques were combined (e.g., structuralism, functionalism, and behaviorism) to deliver the high-quality learning environment; (3) a mixture of face-to-face teacher training and technology-assisted learning (e.g., video, web-based media, movies); and (4) teaching methodologies mixed with actual job duties to integrate studying and working. Blended learning aspired to strike a harmonious and sustainable balance between meaningful physical engagement and online learning by taking into consideration students’ and lecturers’ attitudes and connections (Anthony et al., 2020). Furthermore, because of its many advantages—including time efficiency, the use of state-of-the-art technology, and unconstrained space and time—the blended learning strategy is recognized as advantageous for modern learning. Additionally, it improved students’ learning capacities by providing them with the information and talents they would require once they graduated (Wong et al., 2018). Because it solely concentrates on the ideal mixture for the most suitable educational experience, this approach has consequently continued to be a crucial concept in universities (Ju and Mei, 2018).

In Vietnam, the conference “Application of information technology in blended learning to satisfy the needs of the specific general education curriculum” was established by the Ministry of Education and Training in 2015 which paid attention to the blended learning approach in overall education, and many faculties have been using it to teach a variety of subjects (Dai et al., 2021). According to the Ambient Insight’s (2014) projection, Vietnam has become one of the top ten nations with the greatest self-paced e-learning growth rates in the globe, as well as the Asia area between 2013 and 2018. However, the research on blended learning especially students’ perception and intention in Vietnam has been still limiting in Asia, especially in Vietnam.

The hypothesis development and research framework

The hypothesis development of the research

Bamoallem and Altarteer (2021) found in Saudi Arabia that “Social presence” greatly enhances how students perceive blended learning, which is attributed to elements like emotive expression, open communication, and group cohesiveness. Researchers Zhang and Dang (2020) and Nuamsamrarn and Achwarin (2023) discovered that students’ behavioral intentions and contentment with blended learning were higher when they rated social factors positively. Several researchers have also investigated the favorable benefits of social presence on attitude (Hassanein and Head, 2007) and behavioral intention (Gefen and Straub, 2003) in different circumstances. In the Vietnamese context, existing research has not identified a significant correlation between social presence and students’ inclination towards blended learning. Thus, the author proposed these hypotheses as below:

H1a: Social presence has a positive influence on student perception of blended learning.

H1b: Social presence has a positive influence on behavioral intention towards blended learning.

On the other hand, Bowyer and Chambers (2017) argued that within blended learning, online discussion fosterd a community of inquiry between learners increasing knowledge acquisition. In addition, cognitive presence which was investigated utilizing four distinct criteria including triggering events, explorations, integrations, and resolves, was found to have a favorable impact to students’ perceptions of blended learning (Bamoallem and Altarteer, 2021). Students gained from cognitive presence not just in regards encouragement, but also in terms of progressively adjusting the self-directed learning strategy by using a wide range of data resources (Bao, 2020). Whereas cognitive presence addresses both hard and soft skill learning, social presence fosters an accessible, trustworthy, and dynamic setting in which students can gain the confidence they need to master them (Li, 2015). Additionally, Salloum (2011) also claimed that participants who perceived high cognitive presence and motivation predicted the student’s perception. However, many previous Vietnamese authors have yet to discover a correlation between cognitive presence and student perspective as well as behavioral intention. As a result, this study contained the premise that the cognitive presence has a beneficial influence on student perception and behavioral intention.

H2a: Cognitive Presence has a influence impact on students’ perception about blended learning.

H2b: Cognitive Presence has a positive influence on behavioral intention towards blended learning.

According to Shea et al. (2005), teaching presence was a vital part of the paradigm for the high-quality online learning environment that we were creating as part of our asynchronous learning network study. They contended that students with higher levels of teaching presence, education, and pleasure in courses where their teachers were instructed and encouraged to build effective teaching presence. Cole et al. (2017) investigated the possibility for teacher presence was to impact students’ motivation in distance learning, as well as the ability to predict student motivation toward online courses based on susceptibility to instructor input. Particularly, being knowledgeable of the course books, assigned instructional items, assignments, exercises and additional materials, in particular, functioned as a guideline for pupils, preparing them for the dramatic switch in study mode, which assisted in the presentation of up to 80% of the course’s curriculum (Alshehri et al., 2020). Even though, teaching presence was a predictor of many aspects of students such as their motivation (Cole et al., 2017), their perception (Bamoallem and Altarteer, 2021), no research has found it to affect behavioral intention and student perception in Vietnam blended learning. As a result, the author proposed the following hypothesis:

H3a: Teaching presence has a positive influence on students perception about blended learning.

H3b: Teaching presence has a positive influence on behavioral intention towards blended learning.

The Technology Acceptance Model (TAM) predicts how people will react to new technology, focusing on what influences learners’ decisions to use new learning systems. 1 Specifically, it suggests that students’ learning quality improves when they intend to use an E-learning platform. This intention is driven by their attitude, perceived ease of use, and perceived usefulness. Research has shown that attitude is the most crucial factor in predicting students’ intention to use technology in education, particularly during crises like the Covid-19 pandemic Thus, this study found this relationship in the Vietnamese blended learning context as the following hypothesis:

H4: Student perception has a positive influence on behavioral intention towards blended learning.

As a result of the Covid-19 epidemic, there are a variety of challenges that need to be addressed. Particularly, blended learning in the students’ perception as well as their behavioral intentions was a significant impact. From that, Zhong et al. (2020) argued that a higher understanding of Covid-19 was found to be connected with a lower risk of unfavorable attitudes and potentially dangerous behaviors. Additionally, based on the Theory of Planned Behavior, each individual’s behavioral intent was significantly inspired by their societal community’s ethical norm. During Covid-19 pandemic, the author highly supposed that blended learning meets the mode of online-to-offline learning while satisfying the students’ convenience demands and safety concerns which may shape behavior perception using a blended methodology. Thus, understanding Covid-19 coordinated the relationship between students’ cognitive and behavioral intentions toward blended learning students. The author proposed this hypothesis as follows:

H5: Understanding Covid-19 has a moderating effect between Students’ perceptions and behavioral intentions towards blended learning.

The research framework

The proposed model (Fig. 1) builds on the Community of Inquiry (CoI) framework (Garrison et al., 2000), which emphasizes the importance of social presence, cognitive presence, and teaching presence in blended learning environments. Social presence enhances interaction and connection, contributing to positive student perceptions and intentions (Shea et al., 2005). Additionally, drawing on the Theory of Planned Behavior (Ajzen, 1991), the model posits that students’ perceptions of blended learning affect their behavioral intentions to continue using these methods. The novel inclusion of the Understanding of Covid-19 as a moderating factor reflects how awareness of the pandemic can shape students’ learning preferences and intentions (Bao, 2020).

Fig. 1: The model of The impact of community of inquiry on university students’ behavioral intention towards blended learning.
figure 1

Note: Modified from Bamoallem and Altarteer (2022).

Full size image

Methodology

Methodology

Here’s a revision that addresses the feedback regarding population and sampling technique, as well as sample representativeness:

This study employed a mixed-method approach, combining qualitative and quantitative research. For the quantitative component, a representative sample of 316 students from two universities in Ho Chi Minh City was selected to ensure statistical significance. The sampling technique used was convenience sampling, ensuring that the population of interest, Vietnamese university students, was well-represented. Although convenience sampling has limitations in generalizability, the sample size was sufficiently large to provide meaningful insights into the student population. Data analysis was conducted using SPSS and AMOS software, applying structural equation modeling, descriptive analysis, and a variety of validity and reliability tests. The original English measurement items were translated into Vietnamese with the help of bilingual scholars to ensure cultural and linguistic accuracy. Slight modifications were made based on participant feedback during the pretest phase to improve wording and readability.

Research design

This study utilized an exploratory sequential mixed-methods design, starting with qualitative data collection to refine the survey instrument, followed by quantitative analysis to test the hypotheses. The qualitative phase helped inform and modify the questionnaire items, ensuring they were contextually relevant for Vietnamese students, while the quantitative phase aimed to test relationships between variables using a larger sample. Based on the Community of Inquiry (CoI) framework and existing literature, the research focused on examining the effects of social presence, teaching presence, and cognitive presence on students’ perception and behavioral intention towards blended learning, with Understanding of Covid-19 as a moderating variable. The proposed model is shown in Fig. 1.

The instrument was developed by adapting validated scales from previous research. Items for social presence, cognitive presence, and teaching presence were derived from Garrison et al. (2010) and Bamoallem and Altarteer (2021). These were assessed using a 5-point Likert scale, ranging from “strongly disagree” to “strongly agree.” Social presence was measured with 6 items, cognitive presence with 8 items, and teaching presence with 8 items. The items for student perception and behavioral intention towards blended learning were adapted from Bamoallem and Altarteer (2021), with 3 items per construct. The moderating variable, Understanding of Covid-19, was adapted from Prasetyo et al. (2020) and measured using 4 out of 5 items. To ensure content validity, the instrument underwent a rigorous translation and back-translation process. A bilingual scholar translated the original English items into Vietnamese, and another bilingual researcher translated them back into English. A preliminary assessment was administered to 50 students from two universities in Ho Chi Minh City, who were asked to rate the language, format, logic, and readability of the Vietnamese version. Minor phrasing changes were made based on their suggestions. In addition, a small-scale pretest with ten respondents was undertaken to further enhance the questionnaire by identifying problem items. Experts and academics were also consulted to determine the survey’s suitability and face validity. Based on this feedback, the final instrument’s language and item order were slightly changed.

An initial evaluation was given to 50 students from two universities in Ho Chi Minh City, who were asked to score the language, format, logic, and readability of the Vietnamese version. Minor phrase modifications were implemented based on their ideas. In addition, a small-scale retest of ten respondents was conducted to improve the questionnaire by identifying roblematic items. Researchers and academics were also consulted to evaluate the survey’s suitability and face validity. Based on this feedback, the final instrument’s language and component order were almost altered.

Data collection process

This study was carried out when the country was lockout by Covid-19 pandemic. The schools are closed and students must learn online. Therefore, the online survey was created on Google forms.

The questionnaire was distributed to 350 Vietnamese students in various levels of education who have studied in blended learning. Additionally, the author published an online survey through Facebook, E-mails or Zalo messages which invited them to take part in the online survey. The questionnaire’s URL was shared to answer within a month. Respondents were notified numerous times through online interactions and e-mails.

During the data collection process, ethical considerations were paramount. All participants provided informed consent before participating, and their anonymity and confidentiality were assured throughout the study.

Data specifications

Demographic statistics

There were 350 online surveys in total, of which 316 were acceptable for analyzing; the response rate was 90%, and the data analysis procedure was completed. During the Covid-19 lockout, the author used an online form to gather 316 samples. A total of 146 male and 170 female students participated. People under the age of 20 account for 35%, those aged 20 to 25, 49%, and the remaining 16% are between the ages of 25 and 30 years. Freshmen and seniors both had 20% involvement, with second-year students having the highest percentage at 31%. Finally, the third year yielded 29%. Smartphones and laptops account for 35% and 20% of internet-connected devices in homes nowadays. Desktop computers have the lowest rate, accounting for less than 10%. Thus, based on these findings this official sample size was 316 respondents Table 1.

Table 1 Response rate of groups.
Full size table

Research results

Confirmatory factor analysis (CFA)

The construct loadings were determined using confirmatory factor analysis. CFA goodness-of-fit was used to investigate the constructs’ convergent validity further. CFA had included the following key indexes: Chi-square, Chi- square/df, Index (TLI), Comparative Fit Index (CFI), Root Mean Square Error Approximation (RMSEA). The following indices were calculated: Chi-square/df (cmin/df) = 2.236, Goodness of fit index (GFI) = 0.825, Adjusted goodness of fit index (AGFI) = 0.821, Comparative fit index (CFI) = 0.917 (based on Hair et al. (2009), CFI was greater than 0.90 which was acceptable), Tucker Lewis Index (TLI) = 0.908 (Table 2). According to Hair et al. (2009), RMSEA was equal to or less than 0.08 (RMSEA of this study was 0.065 which was acceptable) (Fig. 2).

Table 2 Confirmatory factor analysis.
Full size table
Fig. 2
figure 2

Confirmatory factor analysis (CFA) was performed toevaluate the measurement model’s overall fit, reliability, convergent validity, and discriminant validity usingAMOS software.

Full size image

Construct validity

The reliability of all the variables including social presence, cognitive presence, teaching presence, student perception and behavioral intention were in the range of 0.820 to 0.936 (Table 3). The results of the reliability tests were acceptable.

Table 3 Fitting Indices for Confirmatory Factor Analysis (CFA).
Full size table

Factor loading of all most items was more than 0.5 (Hair et al., 2009). Firstly, all factors loading were ranging from 0.671 to 0.93, which were high approximately to 0.5. Additionally, each construct’s average variance extracted (AVE) was greater than 0.5, and all latent variables had construct reliability (CR) better than 0.7. Moreover, all the indicators have signification loading into the respective latent constructs with values between 0.537 and 0.780. As a result, the outcomes have been acceptable (Table 3).

To determine the discriminant validity, a comparison should be performed between the value of a factor’s square root of AVE and its inter-construct correlations with other factors with Average Variance Extracted’s square root having a greater value. As a result in Table 3, the square root of Average Variance Extracted (AVE) for each variable in this investigation greater than the inter-construct correlations, showing that the concept has discriminating validity. The last stage of Structural Equation Modeling could begin when the construct validity of the variables had been established.

By comparing the AVE and correlation values, the discriminant validity was determined. discriminant validity was attained when the AVE value for any variable was greater than the squared of correlation between that variable and any other variables. Table 4 demonstrated that the discriminant validity of both samples was adequate and acceptable for all variables. This indicated that the study model satisfied the requirements for discriminant validity validation.

Table 4 Discriminant validity and correlations among the constructs.
Full size table

Hypothesis testing

To investigate the correlation between the model’s constructs and its reliability, structural equation modeling (SEM) was used. This statistical analysis was carried out using the AMOS software. It can be seen from the path diagram that all the criteria for convergent validity have been achieved to a good or even excellent degree: Chi-square/ df = 2.917; df = 343; p = 0.00; GFI = 0.814; CFI = 0.901; RMR = 0.172; RMSEA = 0.078). Normed X2/ df was 1.374 (Bagozzi and Yi, 1988), that the research model achieved the overall fit (Fig. 3).

Fig. 3: The impact of community of inquiry on University Student’s perception of blended learning.
figure 3

The Impact of Community of Inquiry on University Student’s Perception of Blended Learning in Vietnam and The Moderating of “Understanding Covid-19” variable. Extracted from Bamoallem, B., Altarteer, S. Remote emergency learning during Covid-19 and its impact on university students perception of blended learning in KSA. Education Information Technology 27, 157–179 (2022). https://doi.org/10.1007/s10639-021-10660-7.

Full size image

The result supposed that social presence had no significant positive influences student perception (path coefficient = −0.81) and behavioral intention towards blended learning (path coefficient = −0.029), unsupporting H1a and H1b. However, cognitive presence had significant positive influences student perception (path coefficient = 0.112, p < 0.01) and behavioral intention towards blended learning (path coefficient = 0.228, p < 0.001), supporting H2a and H2b. Additionally teaching presence had significant positively influences student perception (path coefficient = 0.373, p < 0.001) and behavioral intention towards blended learning (path coefficient = 0.196, p < 0.001). Thus, H3a and H3b were confirmed. Moreover, student perception had significant positively influences behavioral intention towards blended learning (path coefficient = 0.347, p < 0.001), Thus, H4 was confirmed (Table 5).

Table 5 Results of hypothesis test.
Full size table

The study used the bootstrap method to test the reliability of the model. The results showed that the absolute value of the critical ratio (CR) is less than 2. Thus, we can conclude that the estimate in the linear structural model SEM is reliable (Table 6). All tables now include details on f-squared (effect sizes) and confidence intervals, ensuring that the results are statistically validated and presented with comprehensive analysis.

Table 6 Estimated results using Bootstrap.
Full size table

The f-square index results showed that TP variable has small impact on SPBL variable, CP variable, TP variable and SPBL variable has small impact on BIBL variable (0.02 ≤ f-square < 0.15). SP variable and CP variable has a very small impact on SPBL variable, SP variable has a very small impact on BIBL variable (f-square < 0.02).

Moderation effects

In the Covid-19 epidemic scenario, the moderate impact of “Understanding Covid-19” variable on the linkages between these dependent variables, which were students’ perception and behavioral intention toward blended learning, were estimated using AMOS 22.0 through 316 samples. According to the moderating effect, there was also a significant interaction effect of behavioral intention towards blended learning and understanding of Covid-19 on students’ perception with β = 0.53, p < 0.001, confirming H5. Thus, the findings confirmed our hypothesis for the moderated model. It revealed that students whose understanding of Covid-19 showed more students with a lower level of understanding of Covid-19 had fairly similar student perception of blended learning. Students who had a higher level of understanding Covid-19 reporting higher levels of their perception when having more behavioral intention towards blended learning (Figs. 3 and 4).

Fig. 4: The moderator result.
figure 4

Note: the moderate impact of “Understanding Covid-19” variable on the linkages between these dependent variables, which were students’ perception and behavioral intention toward blended learning.

Full size image

Discussion

The global spread of Covid-19 disrupted the education of over 850 million students globally, leading to widespread school and institution closures (Gopal et al., 2021). In response, authorities adopted virtual learning platforms to maintain access to educational resources (Mahaye, 2020). Blended learning emerged as an effective solution, with Vietnam also adopting this approach. The primary aim of this study was to explore the factors that influence students’ perceptions of blended learning and their behavioral intentions in Vietnamese educational institutions. The research also tested hypotheses based on a modified version of the Community of Inquiry (CoI) framework, focusing on seven hypotheses, including one evaluating the moderating impact of students’ understanding of Covid-19.

The CoI paradigm identifies teaching presence, cognitive presence, and social presence as major characteristics that positively influence students’ impressions of blended learning. However, unlike past research, this study discovered no significant link between social presence and students’ perceptions or behavioral intentions in the Vietnamese environment. In contrast, Bamoallem and Altarteer’s (2021) research found that elements such as design, structure, and facilitation in social presence had a substantial impact on perceptions. However, this study reveals that in Vietnam, stronger social presence does not always translate into enhanced behavioral intention or perception of blended learning.

Prior research has demonstrated the importance of teaching presence in fostering engagement and cognitive presence in online learning environments (Garrison and Arbaugh, 2007;). This study reinforces those findings, revealing that students’ perception of teaching presence positively impacts their behavioral intention toward blended learning. Instructors play an important role in both online and traditional classrooms by encouraging collaboration and providing timely guidance. Cognitive presence, which includes triggering events, investigation, integration, and resolution, has also been highlighted as an important aspect. A stronger cognitive presence not only enhances students’ perceptions but also has a substantial impact on their behavioral intentions in blended learning situations. To improve cognitive presence, teachers should incorporate multimedia tools and instructional videos, resulting in an engaging and dynamic learning environment, especially during hard moments like as the Covid-19 pandemic.

This research also demonstrated that students’ perceptions are a key predictor of their behavioral intention toward blended learning, which is consistent with recent findings from Saudi Arabia (Bamoallem and Altarteer, 2021). The more positively students see blended learning, the greater their behavioral intentions, resulting in a more productive online learning experience. However, online interactions—measured by the frequency of engagement with peers, instructors, and staff—had only a minor impact on perceived learning quality during the emergency transition to remote education. This limited influence may be due to less opportunities for engagement in such environments, despite the continued relevance of these connections for student performance.

A key contribution of this study is the examination of the moderating effect of students’ understanding of Covid-19 on the relationship between their perceptions and behavioral intentions toward blended learning. Students who had a deeper understanding of Covid-19 showed stronger perceptions and behavioral intentions toward blended learning. This finding, which has not been reported in previous research, highlights the unique context of blended learning in Vietnam, especially amid the pressures of the Covid-19 pandemic.

Conclusions

. This study highlights two critical factors from the Community of Inquiry (COI) framework—teaching presence and cognitive presence—that significantly influence students’ perceptions of collaborative learning. These findings align with prior research by Bamoallem and Altarteer (2021), Almalki (2011), and Alshehri et al. (2020). However, unlike Bamoallem and Altarteer’s (2021) research, this study found no significant relationship between social presence and students’ perceptions in the context of blended learning in Vietnam. Additionally, the study addresses a gap in the existing literature by identifying key factors influencing behavioral intention in blended learning. The results confirm that cognitive presence and teaching presence positively impact students’ behavioral intentions. Another notable contribution of this research is its exploration of the moderating role of students’ understanding of Covid-19 in shaping the relationship between their perception of and behavioral intention toward blended learning. The findings suggest that students with a better understanding of Covid-19 are more enthusiastic about blended learning, which enhances their behavioral intentions. These insights provide valuable implications for educational institutions seeking to enhance blended learning adoption by emphasizing cognitive and teaching presence while addressing students’ understanding of external factors such as the pandemic.

Theory and managerial management

The findings of this study have both theoretical and practical implications. Theoretically, this research extends the Community of Inquiry (COI) framework and builds upon prior studies by exploring factors that influence students’ perceptions and behavioral intentions in blended learning. By focusing on the Vietnamese educational context, this study fills a gap in the literature, as no prior research has examined how elements of the COI framework affect students’ perceptions and behavioral intentions toward blended learning in Vietnam. Notably, the study’s results diverge from existing research by identifying no correlation between social presence and students’ perceptions or behavioral intentions in the blended learning context in Vietnam. Additionally, the study highlights the moderating effect of students’ “Understanding of Covid-19” on the relationship between perception and behavioral intention. This unique finding contributes to a deeper understanding of how external factors can influence students’ engagement in blended education.

In addition, the findings of this research have a number of implications. Firstly, the study underscores the importance of the four variables of teaching presence: course design, organization, facilitation, and direct instruction. These elements are critical because students value flexible communication with instructors, which significantly enhances their learning experience (Alshehri et al., 2020). Secondly, academic institutions should prioritize educating students and faculty about Covid-19 measures, such as social distancing and mask-wearing, as well as the associated risks, including infection and mortality rates. Raising awareness about Covid-19 has been shown to have an indirect yet significant impact on perceived learning effectiveness (Prasetyo et al., 2020). By fostering a better understanding of the pandemic, institutions can promote students’ consciousness in collaborative learning and strengthen their behavioral intentions toward adopting the blended learning model.

Limitation and future research

The study’s findings are influenced by certain limitations that offer avenues for future research. First, the geographic focus on schools and universities in Ho Chi Minh City may have shaped the results by reflecting cultural and institutional characteristics unique to the region. This limits the generalizability of the findings to other regions or cultural contexts. To enhance external validity, future studies should collect data from a more diverse and geographically dispersed sample, encompassing educational institutions across Vietnam and even internationally. This broader approach would offer deeper insights into how blended learning is perceived and implemented in varied settings.

Second, while the focus on students’ perceptions and behavioral intentions provides valuable insights, it overlooks other critical perspectives, particularly those of educators. Teachers play a pivotal role in shaping blended learning environments through their attitudes, teaching styles, and interactions with students, all of which significantly influence learning outcomes. Future research should incorporate teachers’ perspectives to offer a more comprehensive understanding of blended learning. Exploring how educators perceive and adapt to blended learning could uncover additional factors essential for its success and sustainability.

Finally, the study relied on the Community of Inquiry (CoI) model, which provided a strong theoretical foundation but limited the exploration of other relevant factors. Future research could expand upon this framework by incorporating variables such as technological readiness, institutional support, and peer interaction. Integrating these elements would offer a more nuanced understanding of blended learning and identify new strategies to enhance student engagement and educational outcomes.