Content area
This study investigated the relations between the five important external factors outlined in the General Extended Technology Acceptance Model for E-Learning and the four core components in Technology Acceptance Model (Perceived Usefulness: PU, Perceived Ease of Use: PEOU, Behavioural Intention of Use: BIU, and Actual Use: AU). It also examined the extent to which these relations differed by academic majors (STEM versus non-STEM). A total of 745 Chinese undergraduates answered a Likert-scale questionnaire. The data were analysed using structural equation modeling, measurement invariance, and multi-group analysis. Except Computer Anxiety, the other four external factors were significantly and positively associated with either PU or PEOU. Although there were no significant differences in the total effects of the five external factors on PU, differences were found in the effects from Computer Self-efficacy, Subjective Norm, and Prior Experience on PEOU between STEM and non-STEM students. Furthermore, the effects of PEOU on PU and BIU were stronger for non-STEM undergraduates; whereas the path from PU to BIU was stronger for STEM undergraduates. The results suggested that different strategies should be implemented according to students’ academic majors to encourage Chinese undergraduates to actively adopt e-learning. For Chinese STEM undergraduates, emphasis should be placed to improve their perceptions of the usefulness aspect in e-learning via strategies such as incorporating gamified elements and engaging features into the learning activities. It is important to enable non-STEM students to operate and navigate the e-learning systems free of effort through essential orientation programs and quality technical support.
Introduction
With the advancement of technology and information systems, the landscape of higher education has undergone a monumental shift towards blended learning designs, marking a pivotal transformation in educational delivery methods worldwide. As traditional classrooms have been increasingly transitioned online, the number of individuals engaging with digital learning platforms surged (Hu & Raman, 2024; Zhai & Ma, 2022). E-learning has often been used interchangeably with other terms, such as computer-based education, computer-aided instruction, and multimedia learning, and computer-assisted learning (Regmi & Jones, 2020).
E-learning offers numerous benefits to student learning. First, e-learning provides effective ways for students to learn as they can choose their preferred time and location to engage with learning, which accommodates university students’ various schedules and life commitments (Al-Balas et al., 2020). Furthermore, e-learning enables students to learn via a wide range of instructional materials, such as audios, videos, emails, animations, and online discussion boards, which not only stimulate learning motivation and but also foster learning engagement. Additionally, e-learning empowers learners to have control over their education, which may promote lifelong learning opportunities for students and enhance their self-directed learning capabilities (Huynh, 2017; Oluwadele et al., 2023).
Because of multiple benefits of e-learning and its wide adoption in higher education, prolific research has been produced to gain an understanding of the underlying factors influencing university students’ adoption of e-learning in the last decades. In recent years, comparative studies have been conducted to examine university students’ adoption of different e-learning systems (Liu et al., 2023); adoption of e-learning in different regions (Jiang et al., 2021) or different countries (Liu et al., 2010). However, comparison of e-learning adoption by students’ academic major has rarely been examined. The present study will contribute to the literature in this aspect by investigating the extent to which university students’ academic major influences their e-learning adoption. Research has indicated that students studying a Science, Technology, Engineering, and Math (STEM) major are more confident about their computer skills and spend more time on computer related activities in their spare time than non-STEM students (Veenstra et al., 2013). The variations in computer related skills and activities may potentially influence students’ acceptance and adoption of e-learning. With the wide spread of e-learning integration into higher education, a nuanced understanding of the impact of academic major (STEM vs. non-STEM) on university students’ e-learning adoption will allow strategies to be designed and implemented by targeting different majors in order to encourage students to actively participate in e-learning.
Literature review
Theoretical models of investigating technology adoption
In the technology adoption literature, Technology Acceptance Model (TAM) has been used as one of the major models to understand the multifaceted factors influencing students’ adoption of learning technologies and information systems (Zhang et al., 2023). The TAM is favoured by researchers to examine factors associated with users’ intension and use of technologies or information systems for two reasons. On the one hand, the TAM is convenient to implement (Abdullah & Ward, 2016; Wu & Zhang, 2014). On the other hand, the TAM has high credibility. King and He (2006) conducted a meta-analysis of 88 studies using the TAM and concluded that the TAM is a highly valid and robust model.
First proposed by Davis (1985), the TAM comprises four core variables, namely Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Behavioral Intention to Use (BIU) and Actual Use (AU) of technologies (Murillo et al., 2021). While PU and PEOU are concerned with user motivation, BIU and AU are the outcome variables on acceptance and adoption (Scherer et al., 2019). PU is conceived as the extent to which an individual perceives that using a technology or an information system would improve the performance in a specific task; whereas PEOU is defined as the extent to which an individual perceives that using a technology or an information system would be effort free (Davis, 1989). PU and PEOU are considered as the two key variables as they affect the two outcome variables directly or indirectly (Scherer et al., 2019). PU and PEOU are said to be influenced by external factors, which can explain variations in the PU and PEOU. The core components in TAM are presented in Fig. 1.
[See PDF for image]
Fig. 1
Core components in the TAM
As argued by Venkatesh and Davis (1996), it is critical to include external factors in the TAM, as without external factors, the TAM lacks capacity to identify the factors underpinning the reasons of users’ adoption of a new technology or information system, but only provides limited information of users’ intention and behaviors of technology usage (Šumak et al., 2011). Understanding what factors and how these factors affect a person’s BIU and AU is especially valuable for improving the design and functionality of novel technologies or information systems as the developers can use such information to improve these aspects (Davis et al., 1989). To address different research foci, researchers have examined a variety of extended versions of TAM by adding a variety of external factors into the original TAM (He et al., 2021). While these extended versions of TAM are suitable for the investigations in specific contexts and populations, they are not able to integrate important external factors to form a unified model of technology adoption.
The GETAMEL
In order to enable the extended versions of TAM to have meaningful applications in understanding usage of learning technologies and e-learning systems, Abdullah and Ward (2016) derived a General Extended Technology Acceptance Model for E-Learning (GETAMEL) by identifying what external factors have received much attention and have been commonly examined between 2006 and 2016. By carefully scrutinizing 107 studies which used the extended versions of TAM 107 studies using the extended versions of TAM, they found altogether there were 152 external factors examined in these studies.
[See PDF for image]
Fig. 2
The paths in the GETAMEL (adapt from Abdullah & Ward, 2016)
Of the 152 external factors, only five had been tested in at least 10 studies, namely Self-efficacy (51 studies), Subjective Norm (32 studies), Perceived Enjoyment (23 studies), Computer Anxiety (19 studies), and Prior Experience (13 studies). Then the average path coefficients (β) were calculated for students, teachers, and other user type respectively with regard to the paths between the five external factors and PU or PEOU. For the student population, all the paths from the five factors to PU were positive: Perceived Enjoyment: β = 0.452, Subjective Norm: β = 0.301, Self-efficacy: β = 0.174, Prior Experience: β = 0.169, and Computer Anxiety: β = 0.070. The paths from four out of five factors to PEOU were positive: Self-efficacy: β = 0.352, Perceived Enjoyment: β = 0.341, Prior Experience: β = 0.221, and Subjective Norm: β = 0.195. The only exception was Computer Anxiety, which negatively predicted PEOU: β = − 0.199. See Fig. 2 for the average paths from the five factors to PU and PEOU.
To test the GETAMEL, Abdullah et al. (2016) examined 242 British computer science undergraduates’ adoption of an e-portfolio system. Their results showed that Prior Experience, Perceived Enjoyment, Self-efficacy, and Subjective Norm all exerted positive predictions of PEOU, whereas only Perceived Enjoyment predicted PU after the contribution made by PEOU. Both PU and PEOU positively predicted BIU, with PU having a much larger effect size than PEOU did.
The present study, research questions, and hypotheses development
The GETAMEL has produced prolific research of more than 100 studies to examine student adoption of various learning technologies and information systems in different learning contexts (Kemp et al., 2019), such as distance learning (Cicha et al., 2021; Jiang et al., 2021), mobile learning applications (Liu et al., 2023), learning management systems (Matarirano et al., 2021), information and communication technologies (Rizun & Strzelecki, 2020), as well as e-learning (Chang et al., 2017; Doleck et al., 2018; Humida et al., 2021). However, there is a lack of research which examines the extent to which students’ academic major may affect their e-learning adoption using the GETAMEL. To the best of the researchers’ knowledge, the only study investigated the influence of academic major on e-learning adoption was conducted by Thongsri et al. (2020). Thongsri et al. found that the paths from Computer Self-efficacy to PU and PEOU differed significantly by academic major (STEM vs. non-STEM). So did the paths from PEOU to PU and from PU to BI. Furthermore, there were also significant differences on the latent mean of Computer Self-efficacy, PEOU, and BI between.
STEM and non-STEM students. STEM students had higher Computer Self-efficacy, perceived that e-learning was easier, and were more likely to adopt e-learning than their non-STEM counterparts.
However, Thongsri et al.’s (2020) study only included one external factor– Computer Self-efficacy in the model, failing to assess how multifaceted factors jointly shape students’ e-learning use (Tamilmani et al., 2021). Additionally, Thongsri et al.’s study only gathered information on students’ intended e-learning adoption rather than their actual e-learning use.
To address these limitations, the present study, conducted amongst Chinese university students, aimed to examine: (1) the extent to which students’ academic major (STEM vs. a non-STEM) influences the relations between the five commonly examined external factors (i.e., Self-efficacy, Subjective Norm, Perceived Enjoyment, Prior Experience, and Computer Anxiety) and the core components in the GETAMEL (i.e., PU, PEOU, BIU, and AU); and (2) if the latent means of the external factors and core components differ by academic major.
Chinese educational system provides a unique context to investigate the impact of academic majors. In China, students need to choose between the two broad academic majors– STEM versus non-STEM– as early as the second year of their high schools in order to determine which subjects they will be examined in the National College Entrance Exam (Gaokao) apart from the three common subjects (i.e., Chinese, mathematics, and English) for both STEM and non-STEM major students in Gaokao. The admission of students in Chinese universities is also based on their majors in high schools. Only students who have studied STEM majors in high schools can study a STEM undergraduate degree at universities. Likewise, only students who have taken a non-STEM major in high schools can be admitted to a non-STEM undergraduate degree (Gu et al., 2018). For some degrees, such as Bachelor of Business, different universities use different criteria to admit students. Some universities only admit students who have studied a non-STEM major in high school, other universities admit students who have studied a STEM major in high school. There are also universities accept.
both STEM and non-STEM major students to study Bachelor of Business. The early division of students by their majors may result in different perceptions, attitudes, and abilities of using computers, learning technologies, and information systems between Chinese STEM and non-STEM students. Such differences may not necessarily affect students’ learning in high schools, as learning and teaching in Chinese high schools are mainly delivered face-to-face (Li et al., 2021). However, as e-learning has been increasingly integrated into the learning and teaching processes and are widely implemented in Chinese universities (Wu, 2024), the early distinction between STEM and non-STEM majors in China is likely to influence students’ intention and adoption of e-learning at tertiary level.
Understanding how the associations between the external factors and the core components (i.e., PU, PEOU, BIU, and AU); and how these constructs may be affected by Chinese university students’ academic major will provide actionable knowledge to university leaders, teachers, and e-learning designers to develop targeted strategies for students with different majors to improve on these factors, which in turn are likely to help Chinese students learn more effectively via e-learning.
Specifically, the study sought to answer three research questions:
1) To what extent do Chinese undergraduates’ e-learning adoption support the hypothesised relations in the GETAMEL? The hypothesised model is displayed in Fig. 3.
According to the GETAMEL, five hypotheses were formulated with regard to the relations between the five external factors and PU:
Hypothesis 1a: Chinese undergraduates’ Computer Self-efficacy is positively associated with their PU of e-learning;
Hypothesis 1b: Chinese undergraduates’ Subjective Norm is positively associated with their PU of e-learning;
Hypothesis 1c: Chinese undergraduates’ Perceived Enjoyment is positively associated with their PU of e-learning;
Hypothesis 1d: Chinese undergraduates’ Prior Experience is positively associated with their PU of e-learning;
Hypothesis 1e: Chinese undergraduates’ Computer Anxiety is positively associated with their PU of e-learning;
Five hypotheses were developed concerning the relations between the five external factors and PEOU:
Hypothesis 2a: Chinese undergraduates’ Computer Self-efficacy is positively associated with their PEOU of e-learning;
Hypothesis 2b: Chinese undergraduates’ Subjective Norm is positively associated with their PEOU of e-learning;
Hypothesis 2c: Chinese undergraduates’ Perceived Enjoyment is positively associated with their PEOU of e-learning;
Hypothesis 2d: Chinese undergraduates’ Prior Experience is positively associated with their PEOU of e-learning;
Hypothesis 2e: Chinese undergraduates’ Computer Anxiety is negatively associated with their PEOU of e-learning;
Four hypotheses were generated for the relations between the four core components in the TAM:
Hypothesis 3a: Chinese undergraduates’ PEOU of e-learning is positively associated with their PU of e-learning;
Hypothesis 3b: Chinese undergraduates’ PU of e-learning is positively associated with their BIU of e-learning;
Hypothesis 3c: Chinese undergraduates’ PEOU of e-learning is positively associated with their BIU of e-learning;
Hypothesis 3d: Chinese undergraduates’ BIU of e-learning positively is positively associated with their AU of e-learning.
2) To what extent do the hypothesised relations in the GETAMEL differ between Chinese STEM and non-STEM undergraduates?
3) To what extent do latent means of the five external factors and the four core components in the GETAMEL differ between Chinese STEM and non-STEM undergraduates?
[See PDF for image]
Fig. 3
The hypothesised model
Materials and methods
Participants recruitment and data collection
Participants were recruited from undergraduates in four public universities in China. In accordance with the ethical requirements, we informed the students that the nature of the participation was voluntary and anonymous. Altogether 745 students (189 males, 555 females, and 1 missing gender information) provided informed consent and participated in the study. Of them, 371 reported studying a STEM degree, and the rest of 374 studied a non-STEM degree. Students in the STEM group would receive either a Bachelor of Science (i.e., chemistry, geography, mathematics, biology, material science, and computer science) or a Bachelor of Engineering (i.e., mechanical engineering, industrial engineering, electrical engineering) upon graduation. Students in the non-STEM group would receive a Bachelor of Arts (i.e., Chinese, English, journalism, and history), a Bachelor of Education (education), and a Bachelor of Business (i.e., business administration and finance). Even though business administration and finance majors involve studying mathematics, the participating universities only admitted students who had studied the non-STEM major in high schools and sat the non-STEM subjects in the National College Entrance Exam to entre the two majors. The universities’ enrolment records also showed that none of the participants studied towards double degrees. The participants’ demographic information is presented in Table 1.
Table 1. Participants’ demographic information
item | category | no. | percentage |
|---|---|---|---|
gender | male | 189 | 25.37% |
female | 555 | 74.50% | |
missing | 1 | 0.13% | |
age | < 18 | 8 | 1.07% |
18 | 122 | 16.38% | |
19 | 312 | 41.88% | |
20 | 252 | 33.83% | |
> 20 | 51 | 6.85% | |
STEM majors (n = 371, 49.80%) | chemistry | 63 | 8.46% |
geography | 16 | 2.15% | |
mathematics | 21 | 2.82% | |
biology | 56 | 7.52% | |
material science | 81 | 10.87% | |
computer science | 27 | 3.62% | |
mechanical engineering | 36 | 4.83% | |
industrial engineering | 43 | 5.77% | |
electrical engineering | 28 | 3.76% | |
non-STEM majors (n = 374, 50.20%) | education | 81 | 10.87% |
Chinese | 66 | 8.86% | |
English | 83 | 9.80% | |
business administration | 44 | 5.91% | |
finance | 13 | 1.74% | |
journalism | 90 | 10.74% | |
history | 17 | 2.28% |
Instrument
The instrument was an anonymous online questionnaire, which consisted of two parts. The first part asked students to fill in demographic information, including gender, age, and major. The second part used 5-point Likert style and had 29 items. Of the 29 items, 28 items were used to measure eight scales (Computer Self-efficacy, Subjective Norm, Perceived Enjoyment, Computer Anxiety, Prior Experience, PU, and PEOU, and BIU) and used “1 = strongly disagree, 5 = strongly agree” as anchors. A single item was used to represent the frequency of students’ AU of e-learning with “1 = never and 5 = a lot” as anchors. All the scales were validated and used in previous studies adopting the TAM (Huang, 2016; Lee et al., 2013; Park et al., 2012; Venkatesh & Bala, 2008; Venkatesh & Davis, 1996).
Computer Self-efficacy: the extent to which a student believes that he/she can use computers to complete a specific task (6 items).
Subjective Norm: the extent to which a student perceives that his/her decision on adopting e-learning is influenced by the attitudes and behaviors of his/her significant others (i.e., teachers, classmates, and friends) (3 items).
Perceived Enjoyment: levels of enjoyable feelings towards using e-learning (3 items).
Computer Anxiety: levels of apprehensive or fearful feelings towards using computers (3 items).
Prior Experience: levels of previous experience of using computers and Internet to learn (3 items).
PU: the extent to which a student perceives that using e-learning enhances his/her performance in learning (3 items).
PEOU: the extent to which a student expects that using e-learning is free of effort (3 items).
BIU: students’ intended decision with regard to e-learning adoption (3 items).
AU: the frequency of students’ e-learning adoption (1 item).
Data analysis
To answer the first research question– examination of the hypothesised model (Fig. 2), we conducted confirmatory factor analyses (CFA) and Structural Equation Modeling (SEM) using the maximum likelihood estimator in Mplus 8.10. We used the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the Tucker-Lewis index (TLI) to determine the fit of the models. We followed Hu and Bentler’s (1999) criteria for CFI and TLI: values of CFI > 0.950 and values of TLI > 0.900 are considered as the acceptable fits to the data. We used Marsh et al.’s (2004) criteria for RMSEA (2004): values of RMSEA < 0.060 reflect good statistical fits.
We also calculated the convergent and discriminant validity for the CFA. The convergent validity was established using three criteria: (1) Cronbach’s alpha > 0.700; (2) scale composite reliability (CR) > 0.700; and (3) average variance extracted (AVE) > 0.500 (Fornell & Larcker, 1981). For satisfactory discriminant validity, the AVE’s square root of a scale needs to be larger than its correlation with other scales (Hair et al., 2017).
To answer the second research question (the extent to which the hypothesised relations in the GETAMEL differ between Chinese STEM and non-STEM undergraduates), we conducted measurement invariance tests using multi-group CFA and SEM analyses (Millsap, 2012). We first constructed the configural multi-group CFA model wherein no constraints were placed on the factor loadings to provide a preliminary picture of the factor structure of the model. Following the configural model, we constrained factor loadings to be equal across the STEM and non-STEM students (factor loadings invariance). Subsequently, we further constrained the item intercepts to be equal across the two groups (intercept invariance). To judge the adequacy of the invariance assumptions, we examined CFI, TLI, and RMSEA and followed the recommended criteria by Cheung and Renswold (2002) and Chen (2007). The invariance is tenable when the change in CFI and TLI is < 0.010 and the increase of RMSEA for the more parsimonious model is < 0.015. Once measurement invariance was established, we conducted a multi-group SEM analysis to examine if the hypothesised relationships between Chinese STEM and non-STEM undergraduates differed.
To provide an answer for the third research question (the extent to which the latent means of the five external factors and the four core components in the GETAMEL differ between Chinese STEM and non-STEM undergraduates), on the basis of multi-group SEM, we used the MODEL CONSTRAINT command to calculate the mean differences of the latent constructs between STEM and non-STEM students. We employed the delta method to estimate the standard error (SE), which equates the effect size of the differences in latent construct means between groups to Cohen’s d. This analytic approach offers a straightforward way to model group differences in latent construct means while also providing an appropriate standard error for testing the statistical significance of these differences.
Results
Descriptive statistics
Descriptive statistics of all the constructs used in the study for the overall sample and for students with STEM and non-STEM majors are presented in Table 2.
Table 2. Descriptive statistics
scale | total group | STEM | non-STEM | |||
|---|---|---|---|---|---|---|
M | SD | M | SD | M | SD | |
Computer Self-efficacy | 3.217 | 0.854 | 3.323 | 0.850 | 3.112 | 0.847 |
3.546 | 0.893 | 3.601 | 0.871 | 3.492 | 0.911 | |
3.609 | 0.909 | 3.628 | 0.928 | 3.591 | 0.891 | |
3.748 | 0.912 | 3.838 | 0.898 | 3.658 | 0.918 | |
3.977 | 1.723 | 4.070 | 2.284 | 3.885 | 0.854 | |
3.859 | 0.846 | 3.903 | 0.877 | 3.816 | 0.813 | |
Subjective Norm | 3.779 | 0.886 | 3.817 | 0.903 | 3.741 | 0.869 |
3.285 | 0.944 | 3.307 | 0.982 | 3.262 | 0.906 | |
3.299 | 0.943 | 3.334 | 0.959 | 3.265 | 0.927 | |
Perceived Enjoyment | 3.207 | 0.949 | 3.270 | 0.931 | 3.144 | 0.963 |
3.126 | 0.919 | 3.218 | 0.878 | 3.035 | 0.950 | |
3.150 | 0.927 | 3.240 | 0.897 | 3.061 | 0.949 | |
Computer Anxiety | 2.792 | 0.996 | 2.663 | 0.985 | 2.920 | 0.993 |
2.764 | 0.979 | 2.677 | 0.971 | 2.850 | 0.979 | |
2.734 | 0.96 | 2.639 | 0.980 | 2.829 | 0.931 | |
Prior Experience | 3.322 | 0.99 | 3.350 | 1.003 | 3.294 | 0.976 |
3.510 | 0.916 | 3.534 | 0.922 | 3.487 | 0.911 | |
3.313 | 0.917 | 3.353 | 0.922 | 3.273 | 0.912 | |
PU | 3.098 | 0.889 | 3.162 | 0.842 | 3.035 | 0.930 |
3.181 | 0.882 | 3.256 | 0.849 | 3.107 | 0.908 | |
3.435 | 0.867 | 3.493 | 0.826 | 3.377 | 0.903 | |
PEOU | 3.537 | 0.839 | 3.563 | 0.827 | 3.511 | 0.850 |
3.200 | 0.930 | 3.261 | 0.900 | 3.139 | 0.956 | |
3.552 | 0.860 | 3.580 | 0.823 | 3.524 | 0.896 | |
3.211 | 0.881 | 3.245 | 0.858 | 3.176 | 0.903 | |
BIU | 3.315 | 0.833 | 3.356 | 0.790 | 3.275 | 0.874 |
3.203 | 0.873 | 3.237 | 0.837 | 3.168 | 0.906 | |
3.179 | 0.891 | 3.205 | 0.877 | 3.152 | 0.906 | |
AU | 2.576 | 0.901 | 2.776 | 0.754 | 2.377 | 0.988 |
Results of research question 1– To what extent do Chinese undergraduates’ e-learning adoption support the hypothesised relations in the GETAMEL?
Results of the CFA
The fit statistics of the CFA conducted based on the full sample are displayed in Table 3, which provided an adequate fit (CFI = 0.947, TLI = 0.937, RMSEA = 0.045).
Table 3. Fit statistics of the total group CFA and SEM
model | description | χ2 | df | CFI | TLI | RMSEA |
|---|---|---|---|---|---|---|
Total group 1 | CFA | 842.95 | 122 | 0.947 | 0.937 | 0.045 |
Total group 2 | SEM | 883.72 | 110 | 0.944 | 0.936 | 0.047 |
The factor loadings and convergent validity of the CFA are presented in Table 4. It shows that all the values of Cronbach’s alpha, CR, and AVE were above the recommendations, suggesting that the convergent validity was adequate.
Table 4. Factor loadings and convergent validity of the CFA
CS | SN | PE | CA | PR | PU | PEOU | BIU | |
|---|---|---|---|---|---|---|---|---|
α | 0.797 | 0.785 | 0.855 | 0.867 | 0.709 | 0.798 | 0.742 | 0.819 |
CR | 0.857 | 0.876 | 0.912 | 0.918 | 0.836 | 0.881 | 0.837 | 0.892 |
AVE | 0.509 | 0.705 | 0.787 | 0.790 | 0.633 | 0.713 | 0.563 | 0.734 |
CS1 | 0.637*** | |||||||
CS2 | 0.794*** | |||||||
CS3 | 0.822*** | |||||||
CS4 | 0.799*** | |||||||
CS5 | 0.455*** | |||||||
CS6 | 0.702*** | |||||||
SN1 | 0.905*** | |||||||
SN2 | 0.873*** | |||||||
SN3 | 0.729*** | |||||||
PE1 | 0.872*** | |||||||
PE2 | 0.918*** | |||||||
PE3 | 0.851*** | |||||||
CA1 | 0.874*** | |||||||
CA2 | 0.915*** | |||||||
CA3 | 0.876*** | |||||||
PR1 | 0.680*** | |||||||
PR2 | 0.818*** | |||||||
PR3 | 0.875*** | |||||||
PU1 | 0.825*** | |||||||
PU2 | 0.886*** | |||||||
PU3 | 0.820*** | |||||||
PEOU1 | 0.747*** | |||||||
PEOU2 | 0.725*** | |||||||
PEOU3 | 0.796*** | |||||||
PEOU4 | 0.731*** | |||||||
BIU1 | 0.837*** | |||||||
BIU2 | 0.858*** | |||||||
BIU3 | 0.875*** |
Notes: ***p <.001. CS = Computer Self-efficacy, SN = Subjective Normal, PE = Perceived Enjoyment, CA = Computer Anxiety, and PR = Prior Experience
The discriminant validity of the CFA was also established as Table 5 demonstrates that all the AVE’s square root of a given scale was greater than its correlation coefficients with other scales.
Table 5. Discriminant validity of the CFA
constructs | CS | SN | PE | CA | PR | PU | PEOU | BIU |
|---|---|---|---|---|---|---|---|---|
CS | (0.713) | |||||||
SN | 0.401*** | (0.839) | ||||||
PE | 0.363*** | 0.498*** | (0.881) | |||||
CA | − 0.132*** | − 0.079* | − 0.217*** | (0.889) | ||||
PR | 0.388*** | 0.389*** | 0.388*** | − 0.128** | (0.795) | |||
PU | 0.350*** | 0.462*** | 0.597*** | − 0.169*** | 0.461*** | (0.844) | ||
PEOU | 0.450*** | 0.360*** | 0.515*** | − 0.132*** | 0.437*** | 0.603*** | (0.750) | |
BIU | 0.352*** | 0.435*** | 0.573*** | − 0.143*** | 0.444*** | 0.607*** | 0.559*** | (0.857) |
AU | 0.156*** | 0.183*** | 0.259*** | − 0.045 | 0.215*** | 0.348*** | 0.257*** | 0.399*** |
Notes: *p <.050, **p <.010, ***p <.001. CS = Computer Self-efficacy, SN = Subjective Normal, PE = Perceived Enjoyment, CA = Computer Anxiety, and PR = Prior Experience
Results of the SEM
The total group SEM also provided a satisfactory fit (CFI = 0.944, TLI = 0.936, RMSEA = 0.047). The path coefficients of the SEM and the hypothesis verification are presented in Table 6 and are visualized in Fig. 4.
Table 6. Path coefficients of the total group SEM and the hypothesis verification
construct | path | construct | β | SE | hypothesis | verification |
|---|---|---|---|---|---|---|
Computer Self-efficacy | → | PU | 0.028 | 0.039 | hypothesis1a | No |
Subjective Norm | → | PU | 0.116** | 0.043 | hypothesis1b | Yes |
Perceived Enjoyment | → | PU | 0.505*** | 0.044 | hypothesis1c | Yes |
Computer Anxiety | → | PU | − 0.015 | 0.034 | hypothesis1d | No |
Prior Experience | → | PU | 0.300*** | 0.043 | hypothesis1e | Yes |
Computer Self-efficacy | → | PEOU | 0.258*** | 0.043 | hypothesis2a | Yes |
Subjective Norm | → | PEOU | − 0.053 | 0.047 | hypothesis2b | No |
Perceived Enjoyment | → | PEOU | 0.457*** | 0.048 | hypothesis2c | Yes |
Computer Anxiety | → | PEOU | 0.041 | 0.037 | hypothesis2d | No |
Prior Experience | → | PEOU | 0.295*** | 0.048 | hypothesis2e | Yes |
PEOU | → | PU | 0.467*** | 0.065 | hypothesis3a | Yes |
PU | → | BIU | 0.547*** | 0.073 | hypothesis3b | Yes |
PEOU | → | BIU | 0.305*** | 0.077 | hypothesis3c | Yes |
BIU | → | AU | 0.448*** | 0.032 | hypothesis3d | Yes |
Notes: *p <.050, **p <.010, ***p <.001
[See PDF for image]
Fig. 4
Path coefficients of the total group SEM (only significant paths are shown)
The predictive effects of the five external factors to PU of e-learning (H1a-H1e)
As seen in Fig. 4, the predictive effects of the five external factors on PU of e-learning contain the direct effects and the indirect effects via PEOU of e-learning. In this study, we are more interested in the total effects of the five external factors on student’s PU of e-learning. As expected, results showed that PU of e-learning was positively and significantly predicted by Perceived Enjoyment (total effect: β = 0.505, SE = 0.044, p <.001), Subjective Norm (total effect: β = 0.116, SE = 0.043, p =.007), and Prior Experience (total effect: β = 0.300, SE = 0.043, p <.001) (see Appendix A for the direct and indirect effects from the five external factors to PU). However, both Computer Self-efficacy (total effect: β = 0.028, SE = 0.039, p =.464) and Computer Anxiety (total effect: β = − 0.015, SE = 0.034, p =.661) were not significantly associated with PU of e-learning. Overall, these findings are partially aligned with our hypotheses.
The predictive effects of the five external factors to PEOU of e-learning (H2a-H2e)
Consistent with our hypotheses, students’ Computer Self-efficacy (β = 0.258, SE = 0.043, p <.001), Perceived Enjoyment (β = 0.457, SE = 0.048, p <.001) and Prior Experience (β = 0.295, SE = 0.048, p <.001) had positive effects on PEOU of e-learning. However, the associations of Subjective Norm (β = − 0.053, SE = 0.047, p =.267) and Computer Anxiety (β = 0.041, SE = 0.037, p =.268) with PEOU of e-learning were not statistically significant.
The relations amongst PU, PEOU, BIU, and AU of e-learning (H3a-H3d)
As hypothesised, the relations amongst PU, PEOU, BIU, and AU of e-learning are closely related. Specifically, PEOU had significant and positive predictive effects on PU (β = 0.467, SE = 0.065, p =.002), and both were positively linked to BIU (PEOU: β = 0.305, SE = 0.077, p =.004; PU: β = 0.547, SE = 0.073, p <.001). Additionally, AU was positively predicted by BIU (β = 0.448, SE = 0.032, p <.001). The substantial and positive relations amongst PU, PEOU, BIU, and AU of e-learning is in line with our hypotheses (H3a-H3d).
Results of research question 2– To what extent do the above hypotheses differ between Chinese STEM and non-STEM undergraduates?
The results of measurement invariance tests are presented in Table 7, the configural (unconstrained) multi-group CFA model provided an adequate level of fit to the data (CFI = 0.931, TLI = 0.919, RMSEA = 0.052). The factor loadings were then subsequently constrained to be equal across the two groups. The changes in model fit were negligible (Multi-group 2 vs. Multi-group 1: ΔCFI = 0.001, ΔTLI = 0.002, ΔRMSEA = 0.002), indicating factor loading invariance. Subsequently, the item intercepts were further constrained to be equal across the two groups, supporting intercept invariance. Again, it results in negligible changes in model fit (Multi-group 3 vs. Multi-group 2: ΔCFI = − 0.001, ΔTLI = 0.002; ΔRMSEA = 0.001). These results suggest that the hypothesised relations and latent construct means between Chinese STEM and non-STEM undergraduates can be directly comparable. Based on the most parsimonious CFA model (Multi-group 3), we examined the hypothesised relations using the SEM framework. The multi-group SEM model provided a good fit to the data (Multi-group 4: CFI = 0.929, TLI = 0.921, RMSEA = 0.056).
Table 7. Fit statistics of the multi-group CFA and SEM
model | description | χ2 | df | CFI | TLI | RMSEA |
|---|---|---|---|---|---|---|
Multi-group 1 | Configural CFA | 1358.44 | 684 | 0.931 | 0.919 | 0.052 |
Multi-group 2 | CFA + factor loadings invariance | 1375.45 | 704 | 0.932 | 0.921 | 0.054 |
Multi-group 3 | CFA + intercept invariance | 1397.13 | 724 | 0.931 | 0.923 | 0.055 |
Multi-group 4 | SEM + intercept invariance | 1423.876 | 728 | 0.929 | 0.921 | 0.056 |
The results of multi-group SEM are displayed in Table 8. Amongst the five external factors, differences were identified in the effects from Computer Self-efficacy, Subjective Norm, and Prior Experience on PEOU of e-learning. Specifically, the effects of Subjective Norm (Δβ = − 0.209, SE = 0.092, p =.023) and Prior Experience (Δβ = − 0.208, SE = 0.097, p =.033) were statistically stronger for non-STEM undergraduates, whereas the effect of Computer Self-efficacy (Δβ = 0.193, SE = 0.089, p =.031) was stronger for STEM undergraduates. However, there were no differences in the total effects of the five external factors on PU of e-learning.
Regarding the relations amongst PU, PEOU, BIU, and AU of e-learning, the effects of PEOU of e-learning on PU (Δβ = − 0.304, SE = 0.155, p =.049) and BIU (Δβ = − 0.488, SE = 0.152, p =.001) were stronger for non-STEM undergraduates. However, PU of e-learning was a stronger predictor of BIU for STEM undergraduates (Δβ = 0.356, SE = 0.157, p =.023). There was no difference in the path from BIU to AU of e-learning between the two groups.
Table 8. Results of multi-group SEM by academic major
H | construct | → | construct | STEM | Non-STEM | Δβ | p | differ |
|---|---|---|---|---|---|---|---|---|
H1a | Computer Self-efficacy | → | PU | 0.056 | − 0.003 | 0.059 | 0.439 | No |
H1b | Subjective Norm | → | PU | 0.088 | 0.119* | − 0.032 | 0.691 | No |
H1c | Perceived Enjoyment | → | PU | 0.480*** | 0.539*** | − 0.060 | 0.525 | No |
H1d | Computer Anxiety | → | PU | − 0.005 | − 0.027 | 0.031 | 0.623 | No |
H1e | Prior Experience | → | PU | 0.324*** | 0.212** | 0.112 | 0.190 | No |
H2a | Computer Self-efficacy | → | PEOU | 0.359*** | 0.184** | 0.193 | 0.031* | Yes |
H2b | Subjective Norm | → | PEOU | − 0.165* | 0.052 | − 0.209 | 0.023* | Yes |
H2c | Perceived Enjoyment | → | PEOU | 0.501*** | 0.408*** | 0.165 | 0.128 | No |
H2d | Computer Anxiety | → | PEOU | 0.030 | 0.030 | − 0.001 | 0.987 | No |
H2e | Prior Experience | → | PEOU | 0.182** | 0.393*** | − 0.208 | 0.033* | Yes |
H3a | PEOU | → | PU | 0.419*** | 0.677*** | − 0.304 | 0.049* | Yes |
H3b | PU | → | BIU | 0.685** | 0.319** | 0.356 | 0.023* | Yes |
H3c | PEOU | → | BIU | 0.106 | 0.576*** | − 0.488 | 0.001** | Yes |
H3d | BIU | → | AU | 0.519*** | 0.400*** | 0.038 | 0.577 | No |
Notes: *p <.050, **p <.010, ***p <.001
Results of research question 3– To what extent do the latent means of the five external factors and the four core components in the GETAMEL differ between Chinese STEM and non-STEM undergraduates?
Based on the most parsimonious CFA model– Multi-group 3 (see Table 5), the latent mean differences of the five external factors and the four core components in the GETAMEL between the two groups were calculated and are displayed in Table 9. STEM undergraduates showed higher levels of Computer Self-Efficacy (d = 0.177, SE = 0.079, p =.026), Perceived Enjoyment (d = 0.205, SE = 0.075, p =.006), PU of e-learning (d = 0.193, SE = 0.079, p =.015), and AU of e-learning (d = 0.405, SE = 0.067, p <.001), whereas non-STEM undergraduates had higher level of Computer Anxiety (d = − 0.252, SE = 0.082, p =.002).
Table 9. The differences of latent construct means between Chinese STEM and non-STEM undergraduates
construct | Cohen’s d | SE |
|---|---|---|
Computer Self-efficacy | 0.177* | 0.079 |
Subjective Norms | 0.071 | 0.081 |
Perceived Enjoyment | 0.205** | 0.075 |
Computer Anxiety | -0.252** | 0.082 |
Prior Experience | 0.100 | 0.085 |
PU | 0.193* | 0.079 |
PEOU | 0.126 | 0.086 |
BIU | 0.092 | 0.077 |
AU | 0.405*** | 0.067 |
Notes: *p <.050, **p <.010, ***p <.001
Discussion
The present study adopted the GETAMEL model (Abdullah & Ward, 2016) to examine the relations between the five commonly investigated external factors and PU, PEOU, BIU and AU of e-learning amongst Chinese undergraduates. It also compared the extent to which these relations differed by students’ academic majors (STEM or non-STEM degrees).
External factors impacting Chinese undergraduates’ PU or PEOU of e-learning
Our study showed that when the five external factors were investigated in a single model, not all the factors were significant contributors to Chinese undergraduates’ PU or PEOU of e-learning. Our results confirmed six out of ten hypotheses regarding the relations between the five external factors and the core components in the GETAME model. We found that two factors– Perceived Enjoyment and Prior Experience significantly predicted both PU and PEOU, indicating that Chinese undergraduates’ enjoyable feelings and positive emotions in using e-learning, or their previous contacts with e-learning systems and applications, are influential not only on how they feel about the usefulness of e-learning systems and applications but also on how they perceive the degree of easiness of using them. Previous research using TAM has also identified that Perceived Enjoyment and Prior Experience are influential factors on students’ PU of metaverse (İbili et al., 2023), flipped classrooms (Ateş et al., 2023), digital technologies (Sprenger & Schwaninger, 2023), and e-portfolio (Abdullah et al., 2016).
Furthermore, consistent with previous findings that Computer Self-efficacy played a significant role of on students’ PEOU of e-portfolio (Abdullah et al., 2016) or intelligent tutoring systems (Han, 2024), our study showed that Computer Self-efficacy also had a significant path to PEOU in the context of e-learning. However, different from our hypothesis, the relation between Computer Self-efficacy and PU was not significant. These results suggested that Chinese students’ confidence about their capabilities to execute tasks using computers were related to the ease aspect rather than the usefulness aspect of e-learning. Our findings are in line with Zhao et al.’s (2021) claim that PEOU tends to be more closely related to an individual’s internal factors, such as one’s ability and competence, rather than external factors, like the influences from other peoples’ opinions. Zhao et al. further argue that external factors tend to be more closely associated with students’ PU. Indeed our study further showed that the influence from students’ significant others (Subjective Norm) was only significantly related to PU rather than PEOU. A number of previous studies also reported that Social Influence was only significantly related to PU in the contexts of learning management system adoption (Binyamin et al., 2018) and use of virtual reality (Liu et al., 2023).
The other two unsupported hypotheses were the non-significant paths from Computer Anxiety to PU and PEOU. Although these findings contradicted some previous studies (Calisir et al., 2014; Chen & Tseng, 2012; Karaali et al., 2011), they were consistent with more recent research (Althubaiti et al., 2022; Matarirano et al., 2021). One possible explanation could be the widespread and ubiquitous use of computers in these days than ten years ago. Nowadays, Chinese university students use computers on a daily basis to fulfill multiple purposes in their lives, such as studying, part-time job, social interaction, and leisure time. The frequent contact with computers makes it unlikely that students experience much anxious feelings of using computers. Hence, Computer Anxiety did not play a significant role in students’ PU and PEOU of e-learning systems.
The relations amongst the core components in the TAM
All the four hypotheses regarding the relations amongst the core components in TAM were results supported. Consistent with TAM and many previous studies (Cao et al., 2020; Zhang et al., 2023), we also found that Chinese undergraduates’ perceptions of the levels of easiness of an e-learning system (PEOU) were substantially and positively associated with their views on the levels of usefulness of that system (PU). This result seemed to suggest that when Chinese undergraduates found an e-learning system is user-friendly, they also tend to consider the system being useful.
Furthermore, we found that while both PU and PEOU were significantly and positively related to BIU, the association was stronger between PU and BIU than that between PEOU and BIU. Such results were logic as users tend to value more the degree to which using an information system would improve their performance rather than the level of effort of using an information system in their decision-making process of system use. Our results and results from previous technology adoption research all support this argument (Huang et al., 2021; Ni & Cheung, 2022; Zhang et al., 2023).
The influence of academic major on the hypothesised relations
Regarding the influence of academic major on the hypothesised relations, we found that none of the relations between the five external factors and PU differed significantly by academic major. However, three out of five external factors (i.e., Computer Self-efficacy, Social Influence, and Prior Experience) were differently associated with PEOU between the two groups. The Chinese STEM undergraduates’ perceptions of the level of easiness of adopting e-learning systems were more strongly related to their evaluation of confidence and efficacy regarding their abilities of using computers. This result echoed with Thognsri et al.’s (2020), which also showed a stronger relation between Computer Self-efficacy and PEOU for STEM students than non-STEM peers.
For Chinese non-STEM undergraduates, their PEOU tended to be affected by their prior experience of using computers and internet-based learning tools. As non-STEM students were less confident about their competence of using computers, they attached a greater role of their past relevant experience when evaluating the usability of e-learning systems. In contrast, Chinese STEM undergraduates’ higher efficacy of using computers to handle different tasks might enable them to quickly navigate through any e-learning system despite lacking prior experience. This might also offer an explanation of the negative association between STEM students’ PEOU and Social Influence. Because of STEM students’ trust of their computer competence, how they perceive an e-learning system was less likely to be influenced by the attitudes of people around them, such as their teachers, classmates, or friends.
Furthermore, the relations amongst PU, PEOU, and BIU also differed by students’ academic major. Specifically, we found that the relations between PEOU and PU, and between PEOU and BIU were stronger amongst Chinese non-STEM undergraduates than amongst STEM undergraduates. These findings also corroborated those reported in Thognsri et al.’s (2020) study. However, different from Thognsri et al.’s results that no significant difference was found on the association between PU and BIU by academic major, our study showed that PU and BIU were more strongly associated amongst STEM undergraduates than amongst non-STEM undergraduates. These results seemed to indicate that in the process of decision making regarding using e-systems or not, the ease of use aspect was valued more by Chinese non-STEM undergraduates, whereas the usefulness aspect was more important to Chinese STEM undergraduates. Hence, to enhance Chinese undergraduates’ intention of adopting e-learning, target strategies should be designed according to students’ academic majors.
The differences of the latent construct means by academic major
Regarding the differences of the latent construct means, significant differences were found on three external factors and two core components in the GETAMEL. Chinese STEM undergraduates scored higher in Computer Self-Efficacy than Chinese non-STEM undergraduates, indicating that Chinese STEM undergraduates were more confident about their abilities of using computers to perform different tasks. Two possible reasons may offer explain this result. First, as students studying towards a STEM degree often need to use many specialised computer software and packages to complete certain tasks in their disciplines, STEM students have more opportunities to use computers in their learning, which may enhance their abilities and confidence in computer use (Das & Bhattacharyya, 2023). Another possibility could be that STEM students’ perceptions of their higher competence in using computers might have prompted them to study a STEM major (Lent et al., 2008). The similar result was also reported by Thongsri et al. (2020), who found Chinese STEM students had higher Computer Self-efficacy than non-STEM students.
However, different from Thongsri et al.’s (2020) study, which found significant differences in Chinese students’ PEOU rather than PU; our study showed the opposite results that significant differences were found on PU rather than PEOU by academic major. Chinese STEM undergraduates perceived e-learning was more useful than Chinese non-STEM undergraduates. Furthermore, STEM students also showed more enjoyable feelings when using e-learning than their non-STEM peers, who showed higher Computer Anxiety than STEM students. These results might also explain more frequent use of e-learning by STEM participants than non-STEM participants in our study.
Conclusions
Practical contributions of the study
Various forms of e-learning have become an integrated part of contemporary higher education across nations (Awan et al., 2021). Helping university students learn more effectively in technology-enhanced learning environments is a critical issue faced by educators, leaders, and institutions around the world. The findings of this study offer some valuable insights of how to encourage Chinese undergraduates to actively adopt e-learning.
As the usefulness and ease of use aspects have different impacts on STEM and non-STEM students’ intention of e-learning adoption, Chinese institutions should consider implementing different strategies according to students’ academic majors. To enhance Chinese STEM undergraduates’ BIU, emphasis should be placed to improve their perceptions of the usefulness of e-learning due to a stronger relation between PU and BIU. One strategy to make STEM students feel e-learning is valuable in their study is to invite senior students to share their experience of how they benefit from e-learning. Senior students may talk about how some of specific STEM learning experiences that would otherwise be impossible have been achieved through e-learning elements. For instance, the use of simulations to replicate real-life clinical scenarios allows medical students to practice and enhance skills safely and effectively (Elendu et al., 2024). Experimenting in virtual laboratories eliminates the risks associated with handling hazardous materials, equipment, or dangerous procedures, ensuring students to achieve practice and learning in a safer environment (Potkonjak et al., 2016).
Helping STEM students to recognise the usefulness of e-learning can also be achieved via setting up online learning communities, which has been demonstrated to have multiple benefits, such as encouraging active participation, fostering engagement, providing support, and building a sense of belonging. Creating discussion boards or using instant messaging chat groups are convenient and effective ways to set up online learning communities (Divjak et al., 2022; Jia et al., 2021). Moreover, as our study also found Perceived Enjoyment is significantly related to PU amongst STEM students, incorporating some gamified elements into e-learning activities and platforms, such as rewards, badges, progress bars, and leaderboards, to create positive and enjoyable learning experiences may also be a useful approach to enhance STEM students’ PU of e-learning (Zhang et al., 2023).
To increase Chinese non-STEM undergraduates’ intention to adopt e-learning, emphasis should be placed on improving their PEOU, which can be achieved by a number of means. First, to make it easy for non-STEM students to operate and navigate the e-learning platform, universities should choose user-friendly e-learning systems and tools with high-quality usability and interface. Furthermore, universities may design some compulsory training or orientation modules for non-STEM students in order to familiarise them with the common features and functionality used in e-learning platforms, helping them navigate more easily and effectively.
Another important approach to make non-STEM students feel easy to use e-learning is to provide a competent technical support team in the schools and departments with non-STEM majors to provide students with timely support and quality services to solve their technical issues (Rotar, 2022). For effective technical support services, universities may consider implementing direct and indirect support simultaneously. Direct technical support can be delivered through various mediums, such as online chat, remote login, help desk, and helpline. Indirect technical support may be achieved by compiling a pool of documents or a dedicated website, which consists of possible technical issues and ways to solve them of various software and learning management systems involved in e-learning (Alshammari et al., 2016).
Limitations and future direction
Despite the insightful findings and some practical implications, our study also suffered from some limitations. First, the five factors in the present study were the commonly examined personal factors in the TAM, future studies may expand on these factors by including additional personal factors, such as levels of digital literacy (He et al., 2021), hedonic motivation (Almogren, 2022), and personal innovativeness (Strzelecki, 2024). Moreover, non-personal factors related to e-learning should also be investigated along with the personal factors, which may reveal some possible interaction effects between personal and non-personal factors on Chinese undergraduates’ e-learning adoption. Some of these non-personal factors worthy of investigations are: facilitating conditions and technical support of e-learning (Gunasinghe et al., 2020), system quality of e-learning (Salloum et al., 2019) and e-learning design (Alshammari et al., 2016). Furthermore, similar to existing studies in the technology adoption literature (Jeong & Kim, 2017), our study only used a self-reported method to assess Chinese undergraduates’ AU of e-learning. Future studies should complement the self-reported measures with objective measures, such as the sensing method, video recordings, or classroom observations to collect more objective and nuanced information on the frequency and duration of Chinese undergraduates’ e-learning behavior in reality (Miller, 2012). Additionally, our sample has imbalanced gender distribution with females accounting for approximately 75%, which may affect the research results as previous research has shown that gender plays a role in users’ technology adoption, such as using artificial intelligence (Zhang et al., 2023). Hence, future research should make efforts to recruit similar numbers of male and female students in order to minimize possible influences of gender on the research results.
Appendix A The direct and indirect effects from the five external factors to PU
external factors | direct effects | indirect effects via PEOU | total effects | |||
|---|---|---|---|---|---|---|
β | SE | β | SE | β | SE | |
Computer Self-efficacy | − 0.092* | 0.042 | 0.120*** | 0.028 | 0.028 | 0.039 |
Subjective Norm | 0.140*** | 0.042 | − 0.025 | 0.022 | 0.116** | 0.043 |
Perceived Enjoyment | 0.291*** | 0.055 | 0.213*** | 0.038 | 0.505*** | 0.044 |
Computer Anxiety | − 0.034 | 0.033 | 0.019 | 0.018 | − 0.015 | 0.034 |
Prior Experience | 0.162*** | 0.048 | 0.138*** | 0.030 | 0.300*** | 0.043 |
Acknowledgements
This work was supported by a grant from the Australian Research Council awarded to Jiesi Guo (DE230100300).
Author contributions
Conceptualization– FH; Data curation– FH; Formal Analysis– FH & JG; Investigation– FH & JG; Methodology– FH & JG, Project administration– FH; Resources– FH & JG; Visualization– FH; Writing– original draft– FH & JG; Writing– review & editing– FH & JG.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability
The datasets generated and/or analysed during the current study are not publicly available due to ethics requirements, but are available from the corresponding author on reasonable request.
Declarations
Generative AI and AI-assisted technologies in the writing process
During the preparation of this work the authors did not use any generative AI and AI-assisted technologies.
Competing interests
The authors have no competing interests to declare.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
Abdullah, F; Ward, R. Developing a general extended technology acceptance model for E-Learning (GETAMEL) by analysing commonly used external factors. Computers in Human Behavior; 2016; 56, pp. 238-256. [DOI: https://dx.doi.org/10.1016/j.chb.2015.11.036]
Abdullah, F; Ward, R; Ahmed, E. Investigating the influence of the most commonly used external variables of TAM on students’ perceived ease of use (PEOU) and perceived usefulness (PU) of e-portfolios. Computers in Human Behavior; 2016; 63, pp. 75-90. [DOI: https://dx.doi.org/10.1016/j.chb.2016.05.014]
Al-Balas, M; Al-Balas, HI; Jaber, HM; Obeidat, K; Al-Balas, H; Aborajooh, EA; Al-Balas, B. Distance learning in clinical medical education amid COVID-19 pandemic in jordan: Current situation, challenges, and perspectives. BMC Medical Education; 2020; 20,
Almogren, A.S. (2022). Art education lecturers’ intention to continue using the blackboard during and after the COVID-19 pandemic: An empirical investigation into the UTAUT and TAM model. Frontiers in Psychology, 13, 944335. https://www.frontiersin.org/articles/10.3389/fpsyg.2022.944335/pdf
Alshammari, SH; Ali, MB; Rosli, MS. The influences of technical support, self-efficacy and instructional design on the usage and acceptance of LMS: A comprehensive review. Turkish Online Journal of Educational Technology-TOJET; 2016; 15,
Althubaiti, A; Tirksstani, JM; Alsehaibany, AA; Aljedani, RS; Mutairii, AM; Alghamdi, NA. Digital transformation in medical education: Factors that influence readiness. Health Informatics Journal; 2022; [DOI: https://dx.doi.org/10.1177/14604582221075554]
Ateş, H; Garzón, J; Lampropoulos, G. Evaluating science teachers’ flipped learning readiness: A GETAMEL approach test. Interactive Learning Environments; 2023; [DOI: https://dx.doi.org/10.1080/10494820.2023.2255232]
Awan, RK; Afshan, G; Memon, AB. Adoption of E-learning at higher education institutions: A systematic literature review. Multidisciplinary Journal for Education Social and Technological Sciences; 2021; 8,
Binyamin, S; Rutter, M; Smith, S. The influence of computer self-efficacy and subjective norms on the students’ use of learning management systems at King Abdulaziz university. International Journal of Information and Education Technology; 2018; 8,
Calisir, F; AltinGumussoy, C; Bayraktaroglu, A; Karaali, D. Predicting the intention to use a web-based learning system: Perceived content quality, anxiety, Perceived system quality, image, and the technology acceptance model. Human Factors and Ergonomics in Manufacturing & Service Industries; 2014; 24,
Cao, W; Fang, Z; Hou, G; Han, M; Xu, X; Dong, J; Zheng, J. The psychological impact of the COVID-19 epidemic on college students in China. Psychiatry Research; 2020; 287, 112934. [DOI: https://dx.doi.org/10.1016/j.psychres.2020.112934]
Chang, CT; Hajiyev, J; Su, C. Examining the students’ behavioural intention to use e-learning in azerbaijan?? The general extended technology acceptance model for e-learning approach. Computers & Education; 2017; 111, pp. 128-143. [DOI: https://dx.doi.org/10.1016/j.compedu.2017.04.010]
Chen, FF. Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal; 2007; 14,
Chen, H; Tseng, H. Factors that influence acceptance of web-based elearning systems for the in-service education of junior high school teachers in Taiwan. Evaluation and Program Planning; 2012; 35,
Cicha, K; Rizun, M; Rutecka, P; Strzelecki, A. Covid-19 and higher education: First-year students’ expectations toward distance learning. Sustainability; 2021; 13,
Das, AR; Bhattacharyya, A. Is STEM a better adaptor than non-STEM groups with online education: An Indian peri-urban experience. Asian Association of Open Universities Journal; 2023; 18,
Davis, F. (1985). A technology acceptance model for empirically testing new end-user information systems: Theory and results. Doctoral dissertation, Massachusetts Institute of Technology.
Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 319–340. https://doi.org/10.2307/249008
Davis, F; Bagozzi, R; Warshaw, P. User acceptance of computer technology: A comparison of two theoretical models. Management Science; 1989; 35, pp. 982-1003. [DOI: https://dx.doi.org/10.1287/mnsc.35.8.982]
Divjak, B; Rienties, B; Iniesto, F; Vondra, P; Žižak, M. Flipped classrooms in higher education during the COVID-19 pandemic: Findings and future research recommendations. International Journal of Educational Technology in Higher Education; 2022; 19,
Doleck, T; Bazelais, P; Lemay, DJ. Is a general extended technology acceptance model for elearning generalizable?. Knowledge Management and E-Learning; 2018; 10,
Elendu, C., Amaechi, D. C., Okatta, A. U., Amaechi, E. C., Elendu, T. C., Ezeh, C. P., & Elendu, I. D. (2024). The impact of simulation-based training in medical education: A review. Medicine, 103(27), e38813. https://doi.10.1097/MD.0000000000038813.
Fornell, C; Larcker, DF. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research; 1981; 18,
Gu, J., Li, X., & Wang, L. (2018). Higher education in China. Springer Singapore.
Gunasinghe, A., Hamid, J.A., Khatibi, A., & Azam, S.F. (2020). The viability of UTAUT-3 in understanding the lecturer’s acceptance and use of virtual learning environments. International Journal of Technology Enhanced Learning, 12(4), 458–481. https://doi.org/10.1504/ijtel.2020.110056
Hair, J; Hult, G; Ringle, C; Sarstedt, M; Thiele, K. Mirror, mirror on the wall: A comparative evaluation of composite-based structural equation modeling methods. Journal of the Academy of Marketing Science; 2017; 45, pp. 616-632. [DOI: https://dx.doi.org/10.1007/s11747-017-0517-x]
Han, F. (2024). Factors affecting Chinese undergraduate medical students’ behavioural intention and actual use of intelligent tutoring systems. Australasian Journal of Educational Technology. https://doi.org/10.14742/ajet.8814
He, T; Huang, Q; Yu, X; Li, S. Exploring students’ digital informal learning: The roles of digital competence and DTPB factors. Behaviour & Information Technology; 2021; 40,
Hu, LT; Bentler, PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal; 1999; 6,
Hu, K; Raman, A. Systematic literature review on the holistic integration of e-learning in universities: Policy, human, financial, and technical perspectives. Contemporary Educational Technology; 2024; 16,
Huang, Y. The factors that predispose students to continuously use cloud services: Social and technological perspectives. Computers & Education; 2016; 97, pp. 86-96. [DOI: https://dx.doi.org/10.1016/j.compedu.2016.02.016]
Huang, H; Chen, Y; Rau, P. Exploring acceptance of intelligent tutoring system with pedagogical agent amongst high school students. Universal Access in the Information Society; 2021; [DOI: https://dx.doi.org/10.1007/s10209-021-00835-x]
Humida, T; Al Mamun, MH; Keikhosrokiani, P. Predicting behavioral intention to use e-learning system: A case-study in begum Rokeya university, rangpur, Bangladesh. Education and Information Technologies; 2022; 27,
Huynh, R. The role of E-learning in medical education. Academic Medicine; 2017; 92,
İbili, E; Ölmez, M; Cihan, A; Bilal, F; İbili, AB; Okumus, N; Billinghurst, M. Investigation of learners’ behavioral intentions to use metaverse learning environment in higher education: A virtual computer laboratory. Interactive Learning Environments; 2023; [DOI: https://dx.doi.org/10.1080/10494820.2023.2240860]
Jeong, H.I., & Kim, Y. (2017). The acceptance of computer technology by teachers in early childhood education. Interactive Learning Environments, 25(4), 496-512. https://doi.org/10.1080/10494820.2016.1143376
Jia, C; Hew, K; Bai, S; Huang, W. Adaptation of a conventional flipped course to an online flipped format during the Covid-19 pandemic: Student learning performance and engagement. Journal of Research on Technology in Education; 2021; [DOI: https://dx.doi.org/10.1080/15391523.2020.1847220]
Jiang, MYC; Jong, MSY; Lau, WWF; Meng, YL; Chai, CS; Chen, M. Validating the general extended technology acceptance model for e-learning: Evidence from an online english as a foreign Language course amid COVID-19. Frontiers in Psychology; 2021; 12, 671615. [DOI: https://dx.doi.org/10.3389/fpsyg.2021.671615]
Karaali, D; Gumussoy, C; Calisir, F. Factors affecting the intention to use a web-based learning system among blue-collar workers in the automotive industry. Computers in Human Behavior; 2011; 27,
Kemp, A; Palmer, E; Strelan, P. A taxonomy of factors affecting attitudes towards educational technologies for use with technology acceptance models. British Journal of Educational Technology; 2019; 50,
King, W., & He, J. (2006). A meta-analysis of the technology acceptance model. Information& Management, 43(6), 740–755. https://doi.org/10.1016/j.im.2006.05.003
Lee, Y; Hsieh, Y; Ma, C. A model of organizational employees’ e-learning systems acceptance. Knowledge-based Systems; 2013; 24,
Lent, RW; Sheu, HB; Singley, D; Schmidt, JA; Schmidt, LC; Gloster, CS. Longitudinal relations of self-efficacy to outcome expectations, interests, and major choice goals in engineering students. Journal of Vocational Behavior; 2008; 73,
Li, F; Jin, T; Edirisingha, P; Zhang, X. School-aged students’ sustainable online learning engagement during covid-19: Community of inquiry in a Chinese secondary education context. Sustainability; 2021; 13,
Liu, I; Chen, M; Sun, Y; Wible, D; Kuo, C. Extending the TAM model to explore the factors that affect intention to use an online learning community. Computers & Education; 2010; 54,
Liu, Y; Sun, JCY; Chen, SK. Comparing technology acceptance of AR-based and 3D map-based mobile library applications: A multigroup SEM analysis. Interactive Learning Environments; 2023; 31,
Marsh, HW; Hau, KT; Wen, Z. In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and bentler’s (1999) findings. Structural Equation Modeling: A Multidisciplinary Journal; 2004; 11,
Matarirano, O; Jere, N; Sibanda, H; Panicker, M. Antecedents of blackboard adoption by lecturers at a South African higher education institution–Extending GETAMEL. International Journal of Emerging Technologies in Learning (iJET); 2021; 16,
Miller, G. (2012). The smartphone psychology manifesto. Perspectives on Psychological Science, 7(3), 221–237. https://doi.org/10.1177/1745691612441215
Millsap, R. E. (2012). Statistical approaches to measurement invariance. Routledge.
Murillo, GG; Novoa-Hernández, P; Rodriguez, RS. Technology acceptance model and moodle: A systematic mapping study. Information Development; 2021; 37,
Ni, A; Cheung, A. Understanding secondary students’ continuance intention to adopt AI-powered intelligent tutoring system for english learning. Education and Information Technologies; 2022; [DOI: https://dx.doi.org/10.1007/s10639-022-11305-z]
Oluwadele, D; Singh, Y; Adeliyi, TT. E-learning performance evaluation in medical education—A bibliometric and visualization analysis. Healthcare; 2023; 11,
Park, Y; Son, H; Kim, C. Investigating the determinants of construction professionals’ acceptance of web-based training: An extension of the technology acceptance model. Automation in Construction; 2012; 22, pp. 377-386. [DOI: https://dx.doi.org/10.1016/j.autcon.2011.09.016]
Potkonjak, V; Gardner, M; Callaghan, V; Mattila, P; Guetl, C; Petrović, VM; Jovanović, K. Virtual laboratories for education in science, technology, and engineering: A review. Computers & Education; 2016; 95, pp. 309-327. [DOI: https://dx.doi.org/10.1016/j.compedu.2016.02.002]
Regmi, K; Jones, L. A systematic review of the factors–enablers and barriers–affecting e-learning in health sciences education. BMC Medical Education; 2020; 20,
Rizun, M; Strzelecki, A. Students’ acceptance of the COVID-19 impact on shifting higher education to distance learning in Poland. International Journal of Environmental Research and Public Health; 2020; 17,
Rotar, O. Online student support: A framework for embedding support interventions into the online learning cycle. Research and Practice in Technology Enhanced Learning; 2022; 17,
Salloum, S. A., Alhamad, A. Q. M., Al-Emran, M., Monem, A. A., & Shaalan, K. (2019). Exploring students’ acceptance of e-learning through the development of a comprehensive technology acceptance model. IEEE access, 7, 128445-128462. https://doi.org/10.1109/ACCESS.2019.2939467
Scherer, R; Siddiq, F; Tondeur, J. The technology acceptance model (TAM): A meta-analytic structural equation modeling approach to explaining teachers’ adoption of digital technology in education. Computers & Education; 2019; 128, pp. 13-35. [DOI: https://dx.doi.org/10.1016/j.compedu.2018.09.009]
Sprenger, DA; Schwaninger, A. Video demonstrations can predict the intention to use digital learning technologies. British Journal of Educational Technology; 2023; 54,
Strzelecki, A. (2024). Students’ acceptance of ChatGPT in higher education: An extended unified theory of acceptance and use of technology. Innovative Higher Education, 49(2), 223-245. https://doi.org/10.1007/s10755-023-09686-1
Šumak, B; Heričko, M; Pušnik, M. A meta-analysis of e-learning technology acceptance: The role of user types and e-learning technology types. Computers in Human Behavior; 2011; 27,
Tamilmani, K; Rana, NP; Wamba, SF; Dwivedi, R. The extended unified theory of acceptance and use of technology (UTAUT2): A systematic literature review and theory evaluation. International Journal of Information Management; 2021; 57, 102269. [DOI: https://dx.doi.org/10.1016/j.ijinfomgt.2020.102269]
Thongsri, N; Shen, L; Bao, Y. Investigating academic major differences in perception of computer self-efficacy and intention toward e-learning adoption in China. Innovations in Education and Teaching International; 2020; 57,
Veenstra, CP; Dey, EL; Herrin, GD. Is modeling of freshman engineering success different from modeling of non-engineering success?. Journal of Engineering Education; 2013; 97, pp. 467-479. [DOI: https://dx.doi.org/10.1002/j.2168-9830.2008.tb00993.x]
Venkatesh, V; Bala, H. Technology acceptance model 3 and a research agenda on interventions. Decision Sciences; 2008; 39,
Venkatesh, V; Davis, F. A model of the antecedents of perceived ease of use: Development and test. Decision Sciences; 1996; 27,
Wu, J. E-learning management systems in higher education: Features of the application at a Chinese vs. European university. Journal of the Knowledge Economy; 2024; [DOI: https://dx.doi.org/10.1007/s13132-024-02159-6]
Zhai, N; Ma, X. Automated writing evaluation (AWE) feedback: A systematic investigation of college students’ acceptance. Computer Assisted Language Learning; 2022; 35,
Zhang, C; Schießl, J; Plößl, L; Hofmann, F; Gläser-Zikuda, M. Acceptance of artificial intelligence amongst pre-service teachers: A multigroup analysis. International Journal of Educational Technology in Higher Education; 2023; 20,
Zhao, Y; Wang, N; Li, Y; Zhou, R; Li, S. Do cultural differences affect users’ e-learning adoption? A meta‐analysis. British Journal of Educational Technology; 2021; 52,
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.