A Systematic Review of the Technology Acceptance Model for the Sustainability of Higher Education during the COVID-19 Pandemic and Identiﬁed Research Gaps

: Over the past three decades, the Technology Acceptance model (TAM) has garnered con-siderable attention in higher education. COVID-19 boosted the development of TAM as multiple studies were rapidly undertaken during the pandemic. This, however, created a gap in our current understanding of the directions and trends of TAM advancement. The purpose of this study is to obtain insight into the advancement of TAM throughout the pandemic. It would assist researchers in comprehending the advancement and direction of TAM studies in higher education, such as gaining an understanding of the prevalent external variables for TAM, the statistical analysis employed, research methodologies, the technologies studied, and the geographic location of the research conducted. Finally, research gaps and future directions for TAM studies are presented. A systematic review utilizing PRISMA was conducted on 104 sampled publications. It was found that self-efﬁcacy, subjective norms, experience, and enjoyment were the external variables most frequently used in TAM, while internal motivation received minimal attention. The existing studies have focused mainly on student samples, so further investigation is needed into lecturers, higher education personnel, and mixed groups. Further study is also required on qualitative and mixed methods, with the partial least square structural equation model currently dominating statistical analysis. Future technologies such as 5G, AI, cloud computing, augmented reality, virtual reality, and BYOD represent new TAM-related research gaps. The majority of studies have been undertaken in Asian countries, such as China and those in southeast Asia. This new systematic literature review provides insight into the trend of TAM advancement in the sustainability of higher education during the pandemic, the identiﬁed research gaps, and recommendations for future research directions. These ﬁndings also serve as a reference for future research by enhancing the foundation established by previous reviews and research on TAM, thereby facilitating the model’s ongoing expansion.


Introduction
Understanding the acceptance and adoption of a technology is crucial in the era of information. The rapid development and deployment of cutting-edge technology have resulted in a more dynamic and organic technology dissemination process than in the past. Without sufficient data on the adoption or rejection of a technology, introducing an invention would be ineffective and a waste of resources. If acceptance factors are ignored during the invention process, it is questionable how a new technology could be improved, and the ways people cognitively process the innovation would be unknown.
prehend their ecosystems and prepare for future improvement. This would open the door for significant workplace innovations that could create an efficient and cost-effective system. The samples employed in higher education-related TAM research must also be examined since this would reveal who is familiar with the innovation that has occurred during this evolution. Understanding the types of samples will help researchers and employers determine how prepared the future workforce is for digitally enhanced organizations.
To further stimulate research in a growing field such as TAM, academics must comprehend the progress and current status of the research undertaken in this area. Providing the scientific community with valuable information, such as the statistical analyses and particular research methods employed, would significantly benefit research endeavors.
Fourth, it is necessary to examine the nature of the studied technologies through which the future of higher education will be shaped. As open distance learning (ODL) has become mainstream, it is crucial to determine which educational tools are required to enable remote education. Employers also demand this information to determine which online collaboration tools and technologies are suited to the digital age workplace. This information would assist researchers and the scientific community in determining the current state of knowledge as well as the existing research gaps that could be the themes of future research initiatives.
Last but not least, the pandemic has been a global catastrophe and phenomenon. Consequently, it is essential that the global community comprehend the TAM research trends in terms of the geographical regions in which the research has been conducted. In an economic sense, this would inform investors as to which global regions are wellprepared for digital innovation and contain digitally ready citizens. Examining all the articles collected and analyzed was expected to enable the authors to formulate TAM research gaps and suggestions for future research into this topic.
To address the stated deficiency, the present new systematic review analyzes the research conducted during the pandemic, focusing on technology acceptance using TAM in higher education. In addition to closing the current research gap, this systematic review analyzes the sampled articles for trends in the use of external variables, the type of samples, statistical analysis, the research methodology, the technologies being studied, and the geographical distribution of the research conducted. Formulating and synthesizing these data using systematic analysis would assist TAM researchers in finding, incorporating, and subsequently developing theoretical and conceptual models that could serve as a foundational model for future research [13] in technology acceptance research in higher education, as Al-Nuaimi and Al-Emran [13] did for learning management systems and TAM. This new systematic review echoes the work of Granić and Marangunić [4], who provided researchers with an overview of TAM-related educational research conducted from 2003 to 2018. Thus, the following research questions are addressed:

1.
What are the external variables used in the studies? 2.
What are the types of samples included in the studies? 3.
What were statistical analysis and research methods employed in the studies? 4.
What are the technologies investigated in the studies? 5.
Where were the studies conducted geographically? 6.
What are the research gaps and directions for future research that can be identified through a systematic literature review?

Previous Reviews
TAM is the most popular of the models introduced by Fred Davis. There is widespread global recognition that the model can be used to comprehend the acceptance of technologies and beyond. For example, it can be used to predict the acceptance of new technologies [14] and identify the reasons why certain technologies are rejected by users [15]. The model has been used in numerous studies in diverse fields, including business [16], healthcare [17], engineering [18], and education [19].
In the context of education, TAM has been employed to conduct empirical studies at the secondary and post-secondary levels. Thus, after more than three decades of extensive research, the model has become the focus of studies within and outside the educational context. The model has been utilized not only in survey or correlational research designs but also by numerous researchers employing systematic review methods. Table 1 presents the previous review studies that have addressed the Technology Acceptance model, as identified in the current review. Publications were produced at the highest frequency in 2014. Taiwan was the most common location for research. The most prevalent model was the original TAM, while the most common analysis was structural equation modeling. Previous literature reviews related to TAM were analyzed systematically. This enabled the researchers to probe the current situation of the topic through a systematic review [36]. In total, 1123 articles were reviewed by 17 research teams. The minimum number of articles included in a review was 18, while the maximum number of articles included was 145. On average, 66.06 articles were reviewed in each research study.
The first review was conducted in 2007 by Yarbrough and Smith [20], and the latest was undertaken in 2022 by Garavand et al. [35]. The reviews covered various topics, with the most frequently reviewed topic being healthcare. Meanwhile, education continued to be missing from the reviewed articles, as the most recent TAM review was written by Granić and Marangunić [4] in 2019, a year before COVID-19. E-learning was studied in 2020 by García-Murillo et al. [33]; however, the impact of COVID-19 was not included in this review as the studies covered were from 2001 to 2019. Five of the most recent reviews only covered studies conducted between 2001 and 2019, with only one review covering studies conducted in 2020. No review of TAM-related papers published during the pandemic has been conducted, demonstrating a knowledge gap.
Despite the extensive and multidimensional synthesis offered by previous reviews, whether in an educational or non-educational context, there is a glaring research gap that has persisted as these reviews were conducted prior to the pandemic. As COVID-19 has revolutionized the landscape of human existence, including education, the significant research undertaken between 2020 and 2022 will be extremely beneficial to the advancement of TAM and human life. It is noticeable that no review related to TAM in education has been published after the pandemic; in contrast, healthcare has received greater attention, with two review articles. This shows a knowledge gap concerning how the educational sector, especially higher education, was able to sustain the impact of COVID-19 from the TAM perspective. Therefore, a new systematic review to bridge this gap is needed.

Methodology
This study represents a systematic literature review of the research published during the COVID-19 pandemic on TAM in higher education. The study was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement and guidelines. The PRISMA statement was created to assist researchers in conducting a systematic review that would allow for a systematic and coherent reporting of the review's purpose, what the authors did, and what they discovered [37,38]. It is a widely accepted set of guidelines for systematic reviews, and it fits the purpose of this new systematic review on TAM perfectly.

Inclusion and Exclusion Criteria
Based on a set of inclusion and exclusion criteria, articles were chosen for inclusion or exclusion from this collection. The established criteria were among the initial criteria used to choose articles for additional analysis. The inclusion and exclusion criteria employed in this study are detailed in Table 2.

Data Sources and Search Strategies
The literature search was conducted between the end of May 2022 and the beginning of June 2022. The search was limited to articles published between January 2020 and two different end periods: 31 May 2022 for the Scopus database and 6 June 2022 for the Web of Science database. The 2020 to 2022 range was selected because the global outbreak of COVID-19 began in 2020; it remains a pandemic at the time of writing. As stated, Scopus and Web of Science were searched as part of this systematic literature review because these two databases cover a wide variety of publishers and are widely recognized globally as reputable indexing bodies.
The search was conducted using specific keywords and search terms based on the logic of Boolean operators. The keywords used for Scopus were Technology Acceptance Model AND Education AND COVID. This resulted in the following search information: (TITLE-ABS-KEY (technology AND acceptance AND model) AND TITLE-ABS-KEY (education) AND TITLE-ABS-KEY (COVID)). The same search terms were used to search the Web of Science database. The search process produced 227 documents from Scopus and 294 documents from Web of Science based on the keywords used. This brought the total number of documents to 521.
The article information from Scopus and Web of Science was then downloaded for additional analysis and comparison to ensure that there was no duplication between the two databases. From the initial 521 documents, 157 were identified as duplicates, reducing the total number of documents to 364.
Then, the documents were further analyzed in detail, based on the stated inclusion and exclusion criteria. Based on these criteria, another 243 documents were excluded. Finally, 104 documents underwent all the identification, screening, and eligibility filtration and were selected for the review analysis. The research flow diagram based on PRISMA guidelines is shown in Figure 1. Involves samples from higher education Studies not using samples from higher education Empirical research Studies that were not empirical, such as reviews Written in English Studies not written in English

Data Sources and Search Strategies
The literature search was conducted between the end of May 2022 and the beginning of June 2022. The search was limited to articles published between January 2020 and two different end periods: 31 May 2022 for the Scopus database and 6 June 2022 for the Web of Science database. The 2020 to 2022 range was selected because the global outbreak of COVID-19 began in 2020; it remains a pandemic at the time of writing. As stated, Scopus and Web of Science were searched as part of this systematic literature review because these two databases cover a wide variety of publishers and are widely recognized globally as reputable indexing bodies.
The search was conducted using specific keywords and search terms based on the logic of Boolean operators. The keywords used for Scopus were Technology Acceptance Model AND Education AND COVID. This resulted in the following search information: (TITLE-ABS-KEY (technology AND acceptance AND model) AND TITLE-ABS-KEY (education) AND TITLE-ABS-KEY (COVID)). The same search terms were used to search the Web of Science database. The search process produced 227 documents from Scopus and 294 documents from Web of Science based on the keywords used. This brought the total number of documents to 521.
The article information from Scopus and Web of Science was then downloaded for additional analysis and comparison to ensure that there was no duplication between the two databases. From the initial 521 documents, 157 were identified as duplicates, reducing the total number of documents to 364.
Then, the documents were further analyzed in detail, based on the stated inclusion and exclusion criteria. Based on these criteria, another 243 documents were excluded. Finally, 104 documents underwent all the identification, screening, and eligibility filtration and were selected for the review analysis. The research flow diagram based on PRISMA guidelines is shown in Figure 1.

Results
The 104 documents were analyzed analytically to answer the formulated research questions in this systematic literature review. The list of articles analyzed in this review is in Appendix A.

External Variables of TAM
The majority of the reviewed articles constituted an expansion of the original TAM. Only 26 studies implemented TAM in its original form, without modification or the incorporation of external variables. To increase the explanatory power of TAM, the remaining reviewed articles incorporated between one and four external variables. The integration of external variables is derived from the criticism that the model would be overly simple if no external variables were included [39,40]. Figure 2 depicts fourteen of the most popular external factors.

Results
The 104 documents were analyzed analytically to answer the formulated research questions in this systematic literature review. The list of articles analyzed in this review is in Appendix A.

External Variables of TAM
The majority of the reviewed articles constituted an expansion of the original TAM. Only 26 studies implemented TAM in its original form, without modification or the incorporation of external variables. To increase the explanatory power of TAM, the remaining reviewed articles incorporated between one and four external variables. The integration of external variables is derived from the criticism that the model would be overly simple if no external variables were included [39,40]. Figure 2 depicts fourteen of the most popular external factors. Self-efficacy and subjective norms were the external variables included most often in the TAM versions identified in this systematic review, followed by enjoyment, experience, and facilitating conditions. The subjective norms variable originated from the Theory of Reasoned Action, developed by Fishbein and Ajzen [41]. Task-Technology Fit and Information System Success (information quality, system quality, and service quality) models were the external models that were integrated most frequently into TAM. Other external factors such as computer anxiety, satisfaction, perceived fear, and social influence also comprised the external variables used to extend TAM. Innovativeness, insecurity, job relevance, optimism, output quality, perceived convenience, perceived playfulness, perceived pleasure, perceived risk, and perceived satisfaction were additional variables that received less attention yet merit mentioning.

Research Samples
Research samples are a crucial component of empirical research, particularly when examining technology acceptance. Previous studies have always incorporated human subjects as research samples [41,42]. Through the current analysis, it was discovered that the studies related to TAM in higher education during the pandemic exhibited several noteworthy patterns, such as the use of various sample types, as shown in Figure 3. The same phenomenon was observed by Granić and Marangunić [4]. Self-efficacy and subjective norms were the external variables included most often in the TAM versions identified in this systematic review, followed by enjoyment, experience, and facilitating conditions. The subjective norms variable originated from the Theory of Reasoned Action, developed by Fishbein and Ajzen [41]. Task-Technology Fit and Information System Success (information quality, system quality, and service quality) models were the external models that were integrated most frequently into TAM. Other external factors such as computer anxiety, satisfaction, perceived fear, and social influence also comprised the external variables used to extend TAM. Innovativeness, insecurity, job relevance, optimism, output quality, perceived convenience, perceived playfulness, perceived pleasure, perceived risk, and perceived satisfaction were additional variables that received less attention yet merit mentioning.

Research Samples
Research samples are a crucial component of empirical research, particularly when examining technology acceptance. Previous studies have always incorporated human subjects as research samples [41,42]. Through the current analysis, it was discovered that the studies related to TAM in higher education during the pandemic exhibited several noteworthy patterns, such as the use of various sample types, as shown in Figure 3. The same phenomenon was observed by Granić and Marangunić [4]. Of all the research samples, 77.88% (81 studies) involved students from universities or colleges in the ecosystem of higher education. Researchers may have wished to assist students in coping with COVID-19-induced stress, leading to a high rate of student participation. Understanding students' perspectives is also essential for universities [43]. Students constitute the largest portion of the higher education community, which explains this result. Students would have borne the brunt of the effects of COVID-19 on higher education, as they would have been confined to their homes, had limited access to campus, and potentially encountered hardware and technical difficulties.
Only 13 research studies involved lecturers as samples; 6 studies involved samples other than students or lecturers. Another four studies combined samples of students and lecturers to better understand technology acceptance by obtaining two different view paradigms. This finding demonstrates that much remains unexplored in relation to lecturers' technology acceptance during COVID-19 based on the TAM theme.
In terms of quantitative research, the largest sample size employed in the reviewed articles was the 1880 university students engaged by Akour et al. [44], followed by Navarro et al. [45], who included 1011 engineering university students; in comparison, 974 samples were used by Sukendro et al. [46]. These studies were conducted in the Asian countries of the United Arab Emirates, the Philippines, and Indonesia. The smallest sample sizes identified from the selected articles were the 20 university students included in the study by Motamedi [47], the 27 university students in the work by Quadir and Zhou [48], and the 50 university students included by Marpaung et al. [49]. This information could serve as a benchmark for future research on TAM sample size requirements for specific studies.

Analysis Technique and Research Approaches
TAM is a widely used framework for analyzing the acceptance of new and existing technologies. In its initial appearance, TAM was synonymous with quantitative approaches. Davis [1] implemented multitrait-multimethod analyses in producing the instrument for the model. Almost a decade later, quantitative approaches continued to be implemented, with the aim of strengthening the model [50]. As the popularity of the structural equation model grew over time, TAM began to be studied with the same analytic method [51,52]. In 2010, mixed-methods research was also employed to better comprehend the model [53]. A year later, Lindsay, Jackson, and Cooke [54]   Of all the research samples, 77.88% (81 studies) involved students from universities or colleges in the ecosystem of higher education. Researchers may have wished to assist students in coping with COVID-19-induced stress, leading to a high rate of student participation. Understanding students' perspectives is also essential for universities [43]. Students constitute the largest portion of the higher education community, which explains this result. Students would have borne the brunt of the effects of COVID-19 on higher education, as they would have been confined to their homes, had limited access to campus, and potentially encountered hardware and technical difficulties.
Only 13 research studies involved lecturers as samples; 6 studies involved samples other than students or lecturers. Another four studies combined samples of students and lecturers to better understand technology acceptance by obtaining two different view paradigms. This finding demonstrates that much remains unexplored in relation to lecturers' technology acceptance during COVID-19 based on the TAM theme.
In terms of quantitative research, the largest sample size employed in the reviewed articles was the 1880 university students engaged by Akour et al. [44], followed by Navarro et al. [45], who included 1011 engineering university students; in comparison, 974 samples were used by Sukendro et al. [46]. These studies were conducted in the Asian countries of the United Arab Emirates, the Philippines, and Indonesia. The smallest sample sizes identified from the selected articles were the 20 university students included in the study by Motamedi [47], the 27 university students in the work by Quadir and Zhou [48], and the 50 university students included by Marpaung et al. [49]. This information could serve as a benchmark for future research on TAM sample size requirements for specific studies.

Analysis Technique and Research Approaches
TAM is a widely used framework for analyzing the acceptance of new and existing technologies. In its initial appearance, TAM was synonymous with quantitative approaches. Davis [1] implemented multitrait-multimethod analyses in producing the instrument for the model. Almost a decade later, quantitative approaches continued to be implemented, with the aim of strengthening the model [50]. As the popularity of the structural equation model grew over time, TAM began to be studied with the same analytic method [51,52]. In 2010, mixed-methods research was also employed to better comprehend the model [53]. A year later, Lindsay, Jackson, and Cooke [54] applied qualitative approaches using ethnographic design to TAM. However, after the pandemic affected higher education, the pattern of analysis techniques and research methods may have changed.
Based on the sampled articles, the systematic analysis utilizing PRISMA identified quantitative approaches as the most common type of research method employed in TAM studies, as illustrated in Figure 4. Of the sampled articles, 93 (89.42%) utilized quantitative methods. Only two articles were qualitative, while nine utilized the mixedmethods paradigm.
affected higher education, the pattern of analysis techniques and research methods may have changed.
Based on the sampled articles, the systematic analysis utilizing PRISMA identified quantitative approaches as the most common type of research method employed in TAM studies, as illustrated in Figure 4. Of the sampled articles, 93 (89.42%) utilized quantitative methods. Only two articles were qualitative, while nine utilized the mixed-methods paradigm.  Based on the sampled articles, the systematic analysis utilizing PRISMA identified quantitative approaches as the most common type of research method employed in TAM studies, as illustrated in Figure 4. Of the sampled articles, 93 (89.42%) utilized quantitative methods. Only two articles were qualitative, while nine utilized the mixed-methods paradigm.

The Technologies Studied
Of the 104 documents included in this review, which covered numerous types of technologies, 34 of them concerned learning management systems or e-learning. This covered 32.69% of all the articles published on TAM in the higher education context during the COVID-19 pandemic. Other technologies that attracted research attention were mobile learning, with eight studies that comprised 7.69% of the total. Technologies related to online teaching and learning activities also gained significant attention. It was discovered that 7.69% of the studies covered online learning (n = 8), 4.81% referred to online teaching (n = 5), and 5.77% examined remote teaching (n = 6). The distribution of the technologies studied is depicted in Figure 6.

The Technologies Studied
Of the 104 documents included in this review, which covered numerous types of technologies, 34 of them concerned learning management systems or e-learning. This covered 32.69% of all the articles published on TAM in the higher education context during the COVID-19 pandemic. Other technologies that attracted research attention were mobile learning, with eight studies that comprised 7.69% of the total. Technologies related to online teaching and learning activities also gained significant attention. It was discovered that 7.69% of the studies covered online learning (n = 8), 4.81% referred to online teaching (n = 5), and 5.77% examined remote teaching (n = 6). The distribution of the technologies studied is depicted in Figure 6. Artificial intelligence, 5G technology, augmented reality, virtual reality, and robotics received less attention. However, these technologies are required to accommodate the effects of Industrial Revolution 4.0, which may present a research opportunity for future scholars.

Geographical Distribution
The majority of the studies examining the acceptance of technologies in higher education based implicitly on TAM were conducted in Asia, with China leading the region with 11 studies. Ten studies were conducted in Indonesia, six in Malaysia and Vietnam, three in Thailand, and two in the Philippines, all within the southeast Asia region, where researchers placed significant emphasis on educational sustainability during the pandemic. In the Middle East, Saudi Arabia (n = 9), the United Arab Emirates (n = 6), Jordan (n = 5), and Iraq (n = 2) dominated the research in this region. The same pattern of research, focused on the Asian context, was observed in south Asia, where six of the sampled articles were from India and three from Pakistan.

Discussion
TAM has been a highly popular model for explaining technology acceptance among users. However, a gap remains in our current knowledge in relation to how the trends and development of the model can ensure the sustainability of higher education during the pandemic. Thus, this systematic literature review aimed to understand how the model had evolved and developed to support the sustainability of higher education. Moreover, Artificial intelligence, 5G technology, augmented reality, virtual reality, and robotics received less attention. However, these technologies are required to accommodate the effects of Industrial Revolution 4.0, which may present a research opportunity for future scholars.

Geographical Distribution
The majority of the studies examining the acceptance of technologies in higher education based implicitly on TAM were conducted in Asia, with China leading the region with 11 studies. Ten studies were conducted in Indonesia, six in Malaysia and Vietnam, three in Thailand, and two in the Philippines, all within the southeast Asia region, where researchers placed significant emphasis on educational sustainability during the pandemic. In the Middle East, Saudi Arabia (n = 9), the United Arab Emirates (n = 6), Jordan (n = 5), and Iraq (n = 2) dominated the research in this region. The same pattern of research, focused on the Asian context, was observed in south Asia, where six of the sampled articles were from India and three from Pakistan.

Discussion
TAM has been a highly popular model for explaining technology acceptance among users. However, a gap remains in our current knowledge in relation to how the trends and development of the model can ensure the sustainability of higher education during the pandemic. Thus, this systematic literature review aimed to understand how the model had evolved and developed to support the sustainability of higher education. Moreover, this review would enable researchers to understand the current TAM knowledge gap through the trends and patterns emerging from this study.

Prevalent External Variables
The articles studied revealed several intriguing discoveries. Self-efficacy, subjective norms, experience, and enjoyment were among the external factors most often included when researchers applied the expanded TAM. Introduced by Albert Bandura, self-efficacy has had a long history of integration with TAM. Among the first to integrate self-efficacy into TAM were Igbaria and Iivari [56]; this is still practiced today. A significant variable, it has remained relevant over time despite being measured using several sample types. Granić and Marangunić [4] comprehensively evaluated the literature before the pandemic, and their findings are consistent with those of our review. Interestingly, the current findings actually resemble those from the 2015 review conducted by the same researchers [57]. It is anticipated that self-efficacy will continue to be applicable to future TAM research for at least another decade, based on the findings of this study and the patterns derived from previous research [4,57]. Self-efficacy is also essential in the workforce of the present day [58]. Researchers are recommended to include self-efficacy in future studies, either in the original model suggested by Bandura [59] or its modified version [39].
The subjective norms variable is the second most commonly used external variable, which agrees with the findings of previous research [57]. This conclusion can be supported because this factor derives from the Theory of Reasoned Action. As TAM is an extension of the Theory of Reasoned Action, this theory undeniably influences the behavioral intentions of users. The theory was also utilized in the unification and expansion of TAM when there was an attempt to create a new model of the Unified Theory of Acceptance and Use of Technology (UTAUT) [60]. Other domains, such as transportation [61], retail [62], and tourism [63], also exhibit the roles played by subjective norms. These findings enable the suggestion that incorporating additional subjective norms into TAM might be fruitful. According to reports during the pandemic, subjective norms have a major predictive influence on the uptake of technology [35,64,65]. Consequently, this factor may be a significant indicator of the current, endemic, and post-pandemic circumstances. As it did for TAM2, TAM3, UTAUT, and UTAUT2, it may have a similar effect on the future evolution of TAM.
Based on the data, experience is also a popular external variable for TAM. It may initially appear that experience was never a significant factor in TAM. However, these findings are not novel because they have been previously reported, albeit with some exclusions [32]. Scherer and Teo [32] determined that prior studies had covered the experience component generically rather than focusing on technology, making it impossible to comprehend the implications. In contrast, the integration of experience into TAM was reported in the literature [57]. Despite this, the absence of a theoretical basis for experience, in contrast to self-efficacy and subjective norms, may hinder the expansion of this variable. Simultaneously, user experience with technology can be connected with the pandemic's impact since activities such as online learning and working from home are improving daily. These activities enhance consumers' knowledge of different forms of technology. Experience may not be the most important factor in future studies, at least for the time being. When analyzing the development of TAM3 in situations unrelated to pandemics, it is possible that experience would become the most important factor [66].
Enjoyment has been added to TAM to supplement the lack of intrinsic drive in the original form, which led to criticism [67]. The current factor of perceived usefulness is overly extrinsic and perhaps ignores the intrinsic drive of consumers. This is an opportunity for a greater examination of TAM by incorporating several perspectives, such as the full spectrum of the Self-Determination Theory [39]. Although not supported by previous studies [26,32], enjoyment is an important aspect of the larger hedonic motivation to accept future technologies [68]. Including hedonic motivation in an acceptance theory such as UTAUT would enhance usefulness, the most accurate predictor of acceptance [69]. Thus, researchers are recommended to investigate enjoyment, either directly or indirectly, through the use of perceived enjoyment perspectives since this could potentially enhance the explanatory power of TAM.
During the pandemic, the Task-Technology Fit and Information System Success models were used to supplement TAM. A common practice among researchers is to utilize model components or the entire model to develop an extended TAM [70][71][72][73]. This methodology was not evident in previous reviews in non-educational contexts [35], but the trend has become more common in educational reviews [13]. Combining TAM with another information technology model, such as Task-Technology Fit, significantly increases the usage variation and enables it to be applied in more context-related circumstances [74]. As Al-Nuaimi and Al-Emran [13] suggested, attention should be paid to the post-adoption stage because one gap in the existing literature concerns how this could drive the integration of the Information System Success model, which centers on the continuous use of technology rather than solely on the usage intention based on the TAM paradigm.

Types of Samples
Research using students as samples is compatible with the existing body of knowledge [4,26]. Compared to the previous reviews, examinations of academics' adoption of technology have gained momentum. Nevertheless, this represents a knowledge vacuum and a significant opportunity for investigation. In addition to the lack of research employing lecturers as samples, an extremely interesting pattern was seen to have emerged through the use of samples comprising both students and lecturers [75][76][77][78]. It would be exciting to pursue this innovation in future research. The second gap that researchers could address is the inclusion of non-academic personnel who belong to the higher education ecosystem as study samples, in addition to students and faculty members. As few scientists have explored this topic, the knowledge of samples other than students and instructors remains limited [79][80][81].

The Prominent Statistical Method and Research Approach
The use of PLS-SEM, the most prominent form of statistical analysis, contrasts with the approaches used in previous reviews [4]. During the pandemic, the deployment of CB-SEM was replaced by that of PLS-SEM, from which it might be deduced that CB-SEM requires more stringent requirements and assumptions than PLS-SEM. During the global COVID-19 outbreak, social distancing was practiced. As a result, researchers had less access to samples. CB-SEM necessitates larger sample sizes than PLS-SEM [55], explaining why researchers preferred PLS-SEM. Second, CB-SEM requires multivariate normal and normally distributed data. PLS-SEM, on the other hand, offers substantial benefits by being less perturbed by non-normal data [82]. Because of the restrictions imposed during the COVID-19 pandemic, researchers had fewer options when selecting samples, and acquiring normally distributed data might have been more difficult. This is supported by the evidence that some reviewed articles also utilized non-parametric statistical analysis [78].
PLS-SEM, CB-SEM, correlation, and regression analyses dominated the statistical analyses conducted, causing other approaches, such as the implementation of the Rasch model [83] and autoethnography [84], to be less clarified. Consequently, it is unclear what contributions these analyses could make to the development of TAM. Therefore, it was theorized that the future application of quantitative analyses other than structural equation models represents a noteworthy research gap and a potential future direction of TAM.
The majority of the quantitatively oriented publications could be explained by the intrinsic character of TAM. It was conceived as a questionnaire-based study with closedended questions, which is synonymous with quantitative research methods [1,50]. As TAM is basically a self-reported measure, qualitative techniques received less consideration. In future directions involving TAM, qualitative paradigms, such as the use of interviews, observations, and data analytics, are proposed. Scherer and Teo [85] made the same proposal of switching to non-self-reporting techniques, an existing research gap in TAM. Thus, qualitative and mixed methods are suggested for future research.

The Technologies Studied
During the COVID-19 pandemic, the explored technologies were crucial to the sustainability of higher education. Evaluating the examined technologies would enable future researchers to comprehend the technologies that were widely used and accepted throughout the pandemic period. As COVID-19 has altered the environments of higher education and the job market, as well as provided more opportunities for online engagement and collaboration, this knowledge would be useful when formulating new industry strategies.
In higher education, learning management systems (LMSs) are the most-researched technologies. Due to the rapid transition from traditional classrooms to online environments, higher education had little time to select an alternative learning platform to the current LMS. The application of LMS has long been practiced, even before the pandemic [86]. Thus, it is one of the most effective and productive options available to universities and colleges. The findings support the previous literature from before and up to the COVID-19 era [4].
Less popular than LMS, pedagogically focused technologies such as online learning [87][88][89][90][91], mobile learning [15,44,73,[92][93][94][95], online education [81,84,96], online teaching [97][98][99][100], and emergency remote teaching [78,101,102] were investigated to a substantial degree. This illustrates an emerging trend to understand the acceptance of tools (such as LMS) rather than approaches. This appears to reflect that TAM, as a technology or tool, should be perceived as useful before being employed by users. The existing research also conceived of tools such as electric vehicles [103] and cloud computing [104]. Likewise, the study of instructional approaches is less common than the study of technologies. This causes a diminished awareness of the scientific community's endorsement of instructional practices. Future studies should address this constraint.
New cutting-edge technologies such as 5G technology [105], artificial intelligence [106], cloud computing [107], augmented reality [108], and virtual reality [108] have also been under-researched. Policies pertaining to ICT integration, such as bring-your-own-device (BYOD) [109] and virtual tours [79], have received little consideration. These technologies are essential for the workplaces of the future [110,111]. As they remain under-researched, future research on these technologies will have a significant impact.

Research Locations
The majority of the chosen papers were from Asia. China may have taken the lead in the research due to its proactive measures and vigorous response to prevent the spread of the virus via its "zero-COVID" policy. Lockdowns in China's metropolitan areas, such as Shanghai, may explain the strong demand for COVID-19-compatible education mechanisms. China's strong economic and financial capacities to allocate research funding to its local university scholars may be a contributory element to this spike. Meanwhile, China has been garnering global interest in the realm of research as China-related research is frequently published. It is expected that research into TAM in China will become more popular. Southeast Asia is home to a number of emerging economies, including Malaysia, Vietnam, Thailand, Indonesia, and the Philippines, which may account for the high number of studies undertaken in that region.

Identified Research Gaps
This systematic review centered on six review perspectives concerning TAM: external variables, sample types, statistical analysis, research approaches, researched technologies, and research locations. Based on a comprehensive analysis of 104 articles, the author posits that several research gaps should be addressed in future studies. The gaps are presented in Table 3. Future studies should include self-efficacy, subjective norms, and experience as these elements have major theoretical roots in TAM and have been included in its enhanced and expanded forms.

Sample types
Regarding academic studies and understanding (among lecturers and college teachers), little is known about the adoption of technologies. Also under-researched is the non-academic workforce in higher education.
There is a lack of mixed-sample research that includes, for instance, both students and faculty members as samples.
More studies using educators and lecturers as samples. TAM should refocus its efforts on sample mixtures.

Statistical analysis and research approaches
Little focus has been placed on implementing qualitative and mixed methods in research involving TAM.
CB-SEM may continue to be applicable in the future, given the growing recognition of the importance of PLS-SEM. The Rasch model should be investigated in the future.

Researched technologies
Compared to technology tools, there is a dearth of research that employs a pedagogical approach.
It is recommended that researchers study future workplace technologies, such as 5G, AI, cloud computing, virtual tours, AR, and VR. Industry-relevant policies, such as bring-your-own-device (BYOD), also require attention.
Research locations Insufficient research from the African, American, and European continents.
It was suggested that research be conducted in the context of China and emerging economies.

Limitations
The database search was conducted towards the end of May and the beginning of June in 2022. Therefore, articles that were indexed after this date may not be included in our review. As the scope of this review is limited to Scopus and Web of Science, which index publications of sufficient quality, this review was unable to include articles not indexed by these databases. Some articles may have been accidentally omitted due to human error; however, we are sure that the vast majority of relevant articles have been covered.

Conclusions
The Technology Acceptance model has attracted substantial attention from researchers worldwide. The unanticipated outbreak of COVID-19 has further accelerated the growth of TAM in other academic domains, especially higher education. Even though several systematic reviews on TAM have been published, the vast majority were written before the crisis. Consequently, there was a deficiency in knowledge regarding the research trends and directions of TAM during the pandemic. As COVID-19 has altered the way the world functions, such information is crucial for global future development.
This new systematic review examined the literature from six research viewpoints, namely, external variables, sample types, statistical analysis, methodological approaches, researched technologies, and research locations. It is determined that TAM will remain relevant to technology acceptance research for an extended period. Self-efficacy, subjective norms, and experience may also remain significant external variables of TAM. The continued integration of models such as Task-Technology Fit and Information System Success models is expected. Future researchers also need to focus on intrinsic motivation.
Future studies must focus on samples other than students, and the use of mixed samples is highly encouraged. Qualitative research would enable a deeper understanding of TAM from users' perspectives, whereas PLS-SEM is anticipated to be the primary quantitative analysis technique. Additional studies on 5G, AI, VR, and BYOD, as well as in locations outside Asia, would contribute to the existing body of knowledge.

Conflicts of Interest:
The authors declare no conflict of interest.