Abstract
Mobile learning (ML) was widely adopted during the coronavirus disease 2019 (COVID-19) pandemic, but its sustained use post-pandemic is not guaranteed. This study identifies the factors influencing university students’ intention to continue using ML. Using the Unified Theory of Acceptance and Use of Technology (UTAUT-2) model, data from 445 students at King Faisal University were analyzed via structural equation modeling. The results showed that attitude toward ML was significantly influenced by effort expectancy (β = 0.620, p < 0.001), performance expectancy (β = 0.521, p < 0.001), and hedonic motivation (β = 0.313, p < 0.001). For continuous intention, habit was the strongest predictor (β = 0.445, p < 0.001), followed by hedonic motivation (β = 0.471, p < 0.001) and attitude (β = 0.175, p < 0.05). Performance expectancy, effort expectancy, social influence, and facilitating conditions had no significant direct effects on continuance intention. These findings confirm habit as the cornerstone of post-pandemic ML continuance, highlighting a shift from utilitarian factors to automated use and enjoyment. Post-pandemic ML integration must strategically foster habitual use and enhance enjoyment, moving beyond utility-focused approaches. This study provides evidence-based insights for educational leaders and platform developers to guide ML’s sustainable integration.
1 Introduction
The successful incorporation of mobile learning (ML) into education requires viewing it as a sustainable process rather than a fleeting initiative (Bachmair and Pachler, 2015). Mainstreaming ML is therefore crucial for achieving a lasting impact on educational quality and access. ML aligns well with the fourth Sustainable Development Goal (SDG 4), which advocates for inclusive and equitable quality education and promotes lifelong learning opportunities for all (UNESCO, 2015). It supports this goal in several ways (Bachmair and Pachler, 2015; Setirek and Tanrikulu, 2015; Afzal and Anwar, 2023). Primarily, ML increases accessibility and inclusivity by leveraging portable devices to help bridge the digital divide. Furthermore, it can offer personalized, self-paced learning experiences that cater to diverse student needs and learning styles. ML also provides opportunities to continue education during disruptions, thereby supporting lifelong learning.
The use of ML in higher education has seen a significant upward trend (Al Mulhem, 2020a). Due to rapid advancements in mobile technology, the majority of universities have integrated online learning tools, including ML platforms (Al Mulhem, 2020b). ML can be defined as a platform where learners access knowledge using mobile devices. During the coronavirus disease 2019 (COVID-19) pandemic, ML became one of the most popular tools for sustaining education in Saudi Arabia (Alsheibani et al., 2019; Almaiah and Al Mulhem, 2020; Chen et al., 2022). It proved to be a vital platform for developing and supporting distance learning for both teachers and students (Ho, 2021; Althunibat et al., 2021; Lutfi, 2022). Previous studies have indicated that ML platforms can improve student learning abilities (Malik et al., 2019). Given these advantages, many researchers have recommended a post-pandemic shift from traditional methods to ML to enhance the learning process (Sandu and Gide, 2019; Ahmad et al., 2022; Chaudhry and Kazim, 2021; Yang, 2022).
This adoption has persisted beyond the crisis. University teachers in Saudi Arabia have continued to use ML in the post-pandemic era, underscoring its increased usefulness and solidifying its role in the educational landscape (Almaiah et al., 2021a). This continued use presents a new challenge for Saudi universities: moving from emergency adoption to the strategic and sustained integration of ML. This shift in focus—from initial adoption to long-term continuance—is reflected in the growing body of post-pandemic ML research (Ermilinda et al., 2024).
While the existing literature provides ample evidence of ML adoption during the COVID-19 pandemic, a more nuanced and critical gap remains in understanding the drivers of its sustained use in the post-pandemic era. Previous research has extensively covered ML usage during the crisis (Almaiah et al., 2021a; Al Mulhem and Almaiah, 2021), its impact on student outcomes at that time (Almaiah et al., 2021b; Al-Maroof et al., 2021a), and its associated challenges (Babatunde et al., 2021). However, the pressing question is no longer about initial adoption but about continuance intention now that traditional learning options have been restored. A new paradigm is emerging from recent international research, which consistently identifies habit as the dominant predictor of technology continuance in the post-pandemic period, from university settings in Indonesia (Ermilinda et al., 2024) to healthcare systems in Ethiopia (Kelkay et al., 2025). While this suggests a global shift, the universality of this “habit-dominance” model cannot be assumed without validation in distinct socio-technical contexts.
This presents a critical research gap: not a simple lack of studies in Saudi Arabia, but a need to test this emerging paradigm within its unique higher education landscape. The Saudi context, characterized by its rapid, state-driven digital transformation and specific cultural dynamics, offers a vital test case. Does habit indeed become the cornerstone of sustained ML use here as well? Furthermore, the shift from mandatory to voluntary use is likely to reshuffle the relative importance of all determinants. It remains unclear which factors from the UTAUT-2 model (e.g., performance expectancy and social influence) diminish in significance and which persist in this new “voluntary” phase. To address this, our study employs the Unified Theory of Acceptance and Use of Technology (UTAUT-2) to investigate post-pandemic ML continuance among Saudi university students. The findings are expected to make two key contributions:
Theoretically, by testing and validating the emerging global consensus on habit as the cornerstone of post-crisis technology continuance within the significant yet under-explored context of Saudi higher education.
Contextually, by quantifying the relative importance of all UTAUT-2 constructs in this new phase, thereby mapping which drivers are most critical for sustaining use when the compulsion of a crisis has passed.
2 Literature review
2.1 The promise and adoption of mobile learning in higher education
Mobile learning (ML) has established itself as a significant component of modern higher education, primarily due to its capacity to enhance learning accessibility and student engagement. The body of research on ML has grown steadily over the past decade (Almaiah and Abdul Jalil, 2014), documenting its well-documented benefits such as time convenience, ease of access, and cost reduction (Hooks et al., 2021). ML actively supports learners by providing on-demand access to educational resources, enabling activities such as downloading materials, submitting assignments, and accessing new knowledge (Al Mulhem and Almaiah, 2021). It serves as a valuable tool for reinforcing classroom instruction (Salloum and Shaalan, 2018) and allows educators to create tailored learning resources (Almaiah et al., 2022a). Consequently, many scholars now advocate for ML platforms to be a core component of modern education (Tahat et al., 2021; Hair et al., 2017), with studies indicating that students often prefer and perform successfully with mobile devices over purely online or traditional methods (Alsyouf et al., 2022; Al-Maroof et al., 2021b). The relevance of ML extends beyond general education, as evidenced by its growing adoption in specialized fields like healthcare training, where its flexibility supports continuing professional development (Kelkay et al., 2025).
2.2 The pandemic as a catalyst and the emergence of a new challenge
The COVID-19 pandemic acted as an unprecedented catalyst, forcing the rapid, large-scale adoption of ML and other online learning tools globally (Al Mulhem, 2020b). In Saudi Arabia, this emergency shift was characterized by the widespread use of platforms to ensure educational continuity (Alturki and Aldraiweesh, 2022). This period generated a significant volume of research, extensively covering ML usage during the crisis (Almaiah et al., 2021a; Al Mulhem and Almaiah, 2021), its impact on student outcomes at that time (Almaiah et al., 2021b), and the considerable challenges faced, including inadequate technical support, poor connectivity, and low digital literacy (Derbali and Ltaifa, 2022; Elumalai et al., 2020).
Despite these challenges, the pandemic period underscored the potential of ML to provide educational flexibility and accessibility (Hassan, 2022), accelerating digital adoption and fostering greater acceptance among educators and students (Alatni et al., 2021). Now, in the post-pandemic era, Saudi universities are building upon these experiences, moving from emergency remote teaching to a more strategic integration of blended learning models (Alqahtani, 2022). This transition—from forced adoption to voluntary, sustained use—is a central theme in contemporary ML research, with studies across different nations highlighting the need to understand “continuance intention” in a world with restored educational choices (Ermilinda et al., 2024). Current efforts focus on improving digital infrastructure, providing training, and ensuring equitable access (Alshathry and Alojail, 2024; Alkabaa, 2022). This transition from forced adoption to voluntary, sustained integration represents a new and critical challenge for educational institutions.
2.3 Identifying the research gap: from adoption to continuance
A synthesis of the existing literature reveals a clear demarcation. While research is abundant on the initial adoption of ML during the pandemic, there is a distinct lack of investigation into the factors that drive its continuous use after the crisis has subsided and traditional learning options are fully available. Previous studies have effectively explained how and why students and teachers started using ML out of necessity (Almaiah et al., 2021c; Al-Maroof et al., 2021a). However, understanding why they would choose to continue using it—a concept known as continuance intention—is a different research question that remains largely unexplored, particularly in the Saudi Arabian context.
This gap is critical because the factors that drive initial adoption in a crisis (e.g., mandatory use, lack of alternatives) may differ significantly from those that foster long-term, sustainable use in a blended learning environment. A recent meta-analysis of digital learning adoption during the pandemic provides a robust baseline, confirming the general predictive power of UTAUT2 constructs but also revealing significant heterogeneity, suggesting that post-pandemic drivers need to be re-examined in specific contexts (Zheng et al., 2025). Emerging empirical studies in the post-pandemic era, such as one conducted in South African universities, highlight that continuance intention is a complex phenomenon often driven more strongly by individual factors such as satisfaction and learning compatibility than by institutional support (Steyn et al., 2024). Therefore, this study identifies a specific research gap: the need to validate the emerging “habit-dominance” paradigm of post-pandemic technology continuance within the Saudi higher education context. While studies from Indonesia (Ermilinda et al., 2024) and Ethiopia (Kelkay et al., 2025) point to habit as the key driver, it is essential to determine if this finding holds true in contexts with different infrastructural, cultural, and educational policies. Furthermore, by applying the full UTAUT-2 model, this research will not only test for the dominance of habit but will also elucidate the complete factor landscape, revealing which other constructs (e.g., Performance Expectancy, Social Influence) retain their influence and which diminish in the transition from mandatory adoption to voluntary continuance.
2.4 Theoretical model and hypothesis development
To explore the factors that explain ML continuance intention from the perspective of Saudi university students in the post-COVID-19 context, a conceptual model is developed based on the Unified Theory of Acceptance and Use of Technology (UTAUT-2) framework. The UTAUT-2 model is a comprehensive technology adoption model that has been widely applied to measure the adoption, acceptance, and use of ML (Urbach and Ahlemann, 2010; Almaiah et al., 2022b; Hair et al., 2022; Lutfi et al., 2022a). It is particularly suited for this investigation as it incorporates Habit as a key construct, allowing us to directly test its purported dominance against other established factors in the post-pandemic era.
As shown in Figure 1, the UTAUT-2 proposed model consists of seven key constructs: performance expectancy, effort expectancy, facilitating conditions, hedonic motivation, social influence, price value, and habit. Prior research has found that UTAUT-2 outperforms the original UTAUT model in explaining variance in usage behaviors (from 40 to 52%) and behavioral intentions (from 56 to 74%) (Chávez Herting et al., 2020). The UTAUT-2 model has also been used to predict students’ behavioral intentions and teachers’ attitudes toward online learning systems (Rudhumbu, 2022). Crucially, the model’s applicability extends to understanding the psychological drivers of long-term use, such as user satisfaction, which is a key outcome of positive UTAUT2 predictors and a direct precursor to continuance intention (García de Blanes Sebastián et al., 2024). Based on these recommendations, the researchers have selected the UTAUT-2 model as the theoretical foundation for the proposed model in this study, as it is expected to provide a more accurate explanation of the factors influencing the intention of learners to continue adopting ML in the post-COVID-19 context. Figure 2 presents the proposed research model, and the subsequent sections will discuss the hypotheses of this study.
Figure 1
The UTAUT-2 proposed model.
Figure 2
The analysis SEM for the proposed model.
2.5 Performance expectancy (PE)
The PE in the ML context refers to the degree to which a learner believes that using an ML platform will improve their learning performance and effectiveness (Al-Emran et al., 2020). When students perceive that ML enhances their capabilities and outcomes, they are more likely to view it favorably. For instance, since learners relied on these tools during the COVID-19 pandemic, their positive experiences with their utility could foster a continued intention to use ML in the post-pandemic era. This is supported by meta-analytic evidence indicating that PE is among the most consistent predictors of behavioral intention across numerous e-learning studies (Zheng et al., 2025). However, in post-pandemic continuance contexts where users have substantial experience, the direct effect of utilitarian factors, such as PE, on intention may be superseded by more experiential factors, such as habit (García de Blanes Sebastián et al., 2024), leaving its continued influence an open question. Hence, the study hypothesizes that the following:
H1: PE has a significant positive influence on students' attitudes toward the ML platform.
H2: PE has a significant positive influence on students’ intention to continue adopting the ML platform.
2.6 Effort expectancy (EE)
The EE refers to the perceived ease of use associated with an ML technology (Salloum et al., 2019). Technologies that are simple to understand and use are adopted more quickly than those that require significant new skills (Lutfi et al., 2022b). This construct aligns with the perceived ease of use in the Technology Acceptance Model (TAM). Consistent with prior studies (Almaiah et al., 2022b; Al-Maroof and Salloum, 2020), the complexity of an application negatively impacts its adoption. The perception of complexity can vary among learners based on their prior experience. However, in contexts where users have gained substantial experience, such as during the pandemic, the direct effect of EE on intention may diminish, giving way to more experiential factors, such as habit (García de Blanes Sebastián et al., 2024). Hence, the study hypothesizes the following:
H3: EE has a significant positive influence on students' attitudes toward the ML platform.
H4: EE has a significant positive influence on students’ intention to continue adopting the ML platform.
2.7 Habit (HB)
Habit is a key construct in UTAUT-2, reflecting the extent to which individuals tend to perform behaviors automatically because of learning (Al-Maroof et al., 2021a). Previous research indicates that habit significantly influences the use of various educational technologies (Al-Maroof and Salloum, 2020; Elareshi et al., 2022), and can transform initial intention into sustained usage. In the context of this study, the extensive use of ML during the pandemic may have formed strong usage habits among students. Recent studies underscore habit’s role as a dominant predictor of continuance intention post-pandemic, sometimes outweighing other factors, across diverse settings from Indonesian universities to Ethiopian healthcare systems (Ermilinda et al., 2024; Kelkay et al., 2025). Therefore, it is proposed that. Crucially, based on the emerging international consensus, we posit that Habit will be the strongest predictor of continuance intention, thereby validating the “habit-dominance” paradigm in the Saudi context:
H5: HB has a significant positive influence on students' attitudes toward the ML platform.
H6: HB has a significant positive influence on students’ intention to continue adopting the ML platform.
2.8 Hedonic motivation (HM)
The HM refers to the pleasure or enjoyment derived from using a technology (Almaiah et al., 2016). In the context of ML, if students find the platform enjoyable or fun to use, it can significantly increase their engagement and intention to use it. Several studies have established that hedonic motivation is a strong predictor of users’ intentions to adopt various educational technologies (Ntsiful et al., 2022). Martins et al. (2018) confirmed its significant role in explaining students’ intention to use ML platforms. Furthermore, HM is not just a driver of initial use but is critically linked to user satisfaction, a key determinant of whether a positive experience translates into long-term continuance (García de Blanes Sebastián et al., 2024). Therefore, the enjoyment gained from using an ML platform is expected to positively influence students’ continuance intention in the post-pandemic era. Hence, the study hypothesizes that the following:
H7: HM has a significant positive influence on students' attitudes toward the ML platform.
H8: HM has a significant positive influence on students’ intention to continue adopting the ML platform.
2.9 Social influence (SI)
The SI refers to the degree to which an individual perceives that important others (e.g., friends and peers) believe they should use a new technology. In an educational context, a student’s decision to adopt an ML platform can be significantly influenced by the opinions and behaviors of their social circle. Previous studies have found that social influence has a substantial impact on the utilization of ML (Nezamdoust et al., 2022), as social support can strengthen the intention to use new technologies (Almaiah et al., 2020). Some studies have even identified social influence as one of the strongest predictors of ML use among learners (Wilson et al., 2021). However, contrasting perspectives emerge in post-pandemic continuance research, where SI’s effect can be non-significant once usage becomes voluntary and individualized, as found in a recent healthcare study (Kelkay et al., 2025). This presents an interesting point of investigation for the Saudi student context. Consequently, this study proposes the following:
H9: SI has a significant positive influence on students’ intention to continue adopting the ML platform.
2.10 Facilitating conditions (FCs)
The FC refers to the degree to which an individual believes that an organizational and technical infrastructure exists to support the use of a system (Taamneh et al., 2022). This includes aspects like reliable internet connectivity, access to necessary devices, and available technical support. Effective FCs are crucial for the smooth utilization of an ML platform. Previous research has consistently reported that facilitating conditions significantly affect users’ intention to use ML (Almaiah et al., 2022d; Kosiba et al., 2022). The critical role of FCs is particularly pronounced in contexts where infrastructural gaps remain, highlighting its potential as a key differentiator for sustainable adoption (Kelkay et al., 2025). Based on this evidence, the study hypothesizes the following:
H10: FC has a significant positive influence on students’ intention to continue adopting the ML platform.
2.11 Attitudes toward ML platform (ATT)
The ATT measures the degree to which a student’s evaluation of the ML platform is favorable or unfavorable. A positive attitude develops when students believe that technology is helpful, beneficial, and effective for their learning. Consistent with prior research in educational technology, attitude has been shown to have a strong positive correlation with behavioral intention (Almaiah et al., 2022e; Alamri et al., 2020a, 2020b). This relationship is reinforced by studies that position attitude as a central mediator, shaped by cognitive evaluations like PE and EE, and ultimately driving the decision to continue using a technology (García de Blanes Sebastián et al., 2024). Therefore, a positive attitude developed through past use is expected to be a critical determinant of students’ continuance intention in the post-pandemic context. Hence, the study hypothesizes the following:
H11: ATT has a significant positive influence on students’ intention to continue adopting the ML platform.
3 Methodology
3.1 Research design
This study employed a quantitative, cross-sectional research design. A survey-based approach was used to collect data, which is appropriate for testing the hypothesized relationships in the proposed UTAUT-2 model (Hair et al., 2017).
3.2 Participants and sampling
The study employed a non-probability, purposive sampling strategy to target students with specific experiences relevant to the research objective. The inclusion criteria were: (1) being a student at King Faisal University (KFU), and (2) having prior experience using the Blackboard mobile learning platform during the COVID-19 pandemic.
After obtaining ethical approval (approval reference number: KFU-REC-2024-JUL-ETHICS2446), data were collected from 445 students who met the criteria. The researchers coordinated with instructors from the College of Computer Science and Information Technology (CCSIT) to distribute the online questionnaire via email. The voluntary participation rate was high. The demographic profile of the participants is presented in Table 1.
Table 1
| Category | Frequency | Percentage (%) | |
|---|---|---|---|
| Gender | Female | 225 | 50.5 |
| Male | 200 | 49.5 | |
| Age | Between 18 and 29 | 381 | 85.6 |
| Between 30 and 39 | 50 | 11.2 | |
| Between 40 and 49 | 14 | 3.2 | |
| Education level | Bachelor | 355 | 79.8 |
| Master | 90 | 20.2 | |
| Doctorate | 0 | 0 | |
The demographic data.
3.3 Measures and instrument validation
The survey instrument was designed to measure the constructs of the adapted UTAUT-2 model. All measurement items were adapted from established scales in the literature to ensure content validity. A summary of constructs, sources, and the number of items is provided in Table 2. All items used a 5-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree), chosen for its suitability in attitude research and reduced respondent fatigue (Almaiah and Nasereddin, 2020). The questionnaire was structured into three sections: demographics, ML usage patterns, and perceptions of UTAUT-2 constructs.
Table 2
| Items | Source | |
|---|---|---|
| Performance expectancy | ||
| PE1 | I continue to use mobile learning in post the COVID-19 because it helps me understand learning materials | Venkatesh et al. (2003), Al-Emran et al. (2020), Alamri et al. (2020b) |
| PE2 | I continue to use mobile learning in post the COVID-19 because it improves in performing my learning activities | |
| Effort expectancy | ||
| EE1 | I do not need much effort when using mobile learning to learn | Venkatesh et al. (2003) and Alamri et al. (2020b) |
| EE2 | I assume learning through mobile learning is easy | |
| EE3 | mobile learning is friendly | |
| Social influence | ||
| SI1 | My teacher advised me to use mobile learning to study my course in post the COVID-19 | Venkatesh et al. (2003), Venkatesh et al. (2012), Wilson et al. (2021) |
| SI2 | My friends use mobile learning to study in post the COVID-19 | |
| Facilitating condition | ||
| FC1 | I have smartphone to learn through mobile learning application | Venkatesh et al. (2003) |
| FC2 | My friends are helping me when I do not know how to use mobile learning to study in post the COVID-19. | |
| Hedonic motivation | ||
| HM1 | I continue to use mobile learning to study my courses in post the COVID-19 because it is fun tool | Venkatesh et al. (2012) |
| HM2 | I continue to use mobile learning to study courses in post the COVID-19 because it is entertaining tool | |
| HM3 | I continue to use mobile learning to study my courses in post the COVID-19 because it is so much fun | |
| Continuous intention | ||
| CI1 | I continue to use mobile learning to learn my courses in post the COVID-19 | Arpaci and Basol (2020) and Rohan et al. (2021) |
| CI2 | I recommend the mobile learning to learn my courses to my friends | |
| Habits | ||
| HB1 | I continue to use mobile learning to learn my courses in post the COVID-19 because I am used to it | Martins et al. (2018), Zacharis and Nikolopoulou (2022), Elareshi et al. (2022) |
| HB2 | I continue to use mobile learning to learn in post the COVID-19 because I am used to repeating recordings | |
| HB3 | I continue to use mobile learning to study my courses in post the COVID-19 because I am used to using it to do my homework | |
| Attitude | ||
| ATT1 | Learning using mobile learning in post the COVID-19 is a good idea | Smeda et al. (2018) and Agustyaningrum et al. (2021) |
| ATT2 | Learning using mobile learning in post the COVID-19 is very interesting for me | |
| ATT3 | Learning using mobile learning in post the COVID-19 is so much fun | |
Items of the study.
To ensure accuracy for the Saudi context, a rigorous translation and validation process was followed:
Translation and Back-Translation: The instrument was translated from English to Arabic by bilingual experts and then back-translated by an independent translator to ensure conceptual equivalence.
Expert Review: The Arabic version was reviewed for clarity and face validity by three professors specializing in information systems.
Pilot Study and Reliability: A pilot study (n = 30) was conducted. The internal consistency of all constructs, measured by Cronbach’s alpha in SPSS v23, exceeded the 0.70 threshold, indicating acceptable reliability (Nunnally, 1978). In the main study, all constructs demonstrated high reliability, with alpha values exceeding 0.80 (see Table 3).
Table 3
| Construct | Cronbach’s alpha |
|---|---|
| PE | 0.881 |
| EE | 0.842 |
| SI | 0.843 |
| FC | 0.790 |
| HM | 0.852 |
| CI | 0.843 |
| HB | 0.896 |
| AT | 0.901 |
Cronbach’s alpha values for the pilot study (Cronbach’s alpha ≥ 0.70).
3.4 Data analysis plan
The data analysis followed a two-stage approach for Structural Equation Modeling (SEM) using partial least squares (PLS-SEM) in SmartPLS software (
Ringle et al., 2015), which is suitable for complex models and prediction-oriented research (
Hair et al., 2017).
Measurement Model Assessment: The reliability and validity of the constructs were evaluated. Internal consistency was confirmed with Cronbach’s alpha and composite reliability. Convergent validity was established by ensuring Average Variance Extracted (AVE) values were above 0.50. Discriminant validity was assessed using the Fornell–Larcker criterion, whereby the square root of the AVE for each construct must be greater than its correlation with any other construct.
Structural Model Assessment: The structural model and hypotheses were tested by examining the path coefficients (β) and their significance levels using a bootstrapping procedure (5,000 subsamples). The model’s explanatory power was also evaluated using the R2 values.
4 Data analysis and results
4.1 Assessment of the measurement model
The results of the measurement model assessment confirmed the reliability and validity of the constructs. As shown in Table 3, all Cronbach’s alpha values and composite reliability values exceeded the recommended threshold of 0.70, indicating high internal consistency. Convergent validity was established, as all AVE values were above 0.50. Furthermore, as shown in Table 4, the square root of each construct’s AVE exceeded its correlations with all other constructs, confirming discriminant validity (Fornell and Larcker, 1981).
Table 4
| FC | SI | HM | HB | EE | PE | ATT | CI | |
|---|---|---|---|---|---|---|---|---|
| FC | 0.370 | |||||||
| SI | 0.400 | 0.375 | ||||||
| HM | 0.543 | 0.309 | 0.325 | |||||
| HB | 0.475 | 0.343 | 0.383 | 0.673 | ||||
| EE | 0.409 | 0.366 | 0.369 | 0.507 | 0.554 | |||
| PE | 0.275 | 0.293 | 0.311 | 0.384 | 0.360 | 0.322 | ||
| ATT | 0.481 | 0.233 | 0.254 | 0.322 | 0.316 | 0.249 | 0.530 | |
| CI | 0.425 | 0.332 | 0.356 | 0.509 | 0.449 | 0.362 | 0.282 | 0.587 |
Discriminant validity test: square root of AVE (on the diagonal) and construct correlations (below the diagonal).
The bold values on the diagonal represent the square root of the Average Variance Extracted (AVE). For discriminant validity, these values should be greater than the off-diagonal correlations in the corresponding rows and columns.
4.2 Discriminant validity
To assess the discriminant validity of each construct, the researchers calculated the square roots of the Average Variance Extracted (AVE) values and compared them with the correlation coefficients among the constructs. This showed that DV is accepted when the results exceed the correlations for each construct. In this case, the square roots of the AVE values exceeded the correlation coefficients for all constructs (Alksasbeh et al., 2019), as shown in Table 4.
4.3 Hypothesis testing and structural model
The results of the hypothesis tests are presented in Table 5 and Figure 2. The analysis revealed a complex pattern of relationships that illuminates the drivers of post-pandemic ML continuance.
Table 5
| Hypothesis | Relationship | Estimate | S. E. | C. R. | p-value | Interpretation | |
|---|---|---|---|---|---|---|---|
| 0.05 | 0.01 | ||||||
| H1 | Performance expectancy → attitude | 0.521 | 0.058 | 8.983 | 0.000 | Significant | Significant |
| H2 | Effort expectancy → attitude | 0.620 | 0.162 | 3.827 | 0.000 | Significant | Significant |
| H3 | Habit → attitude | 0.138 | 0.106 | 1.301 | 0.193 | Not Significant | Not Significant |
| H4 | Hedonic motivation → attitude | 0.313 | 0.079 | 3.962 | 0.000 | Significant | Significant |
| H5 | Performance expectancy → Continuous intention | 0.028 | 0.060 | 0.462 | 0.644 | Not Significant | Not Significant |
| H6 | Effort expectancy → Continuous intention | 0.063 | 0.620 | 0.102 | 0.919 | Not Significant | Not Significant |
| H7 | Habit → Continuous intention | 0.445 | 0.114 | 3.904 | 0.000 | Significant | Significant |
| H8 | Hedonic motivation → Continuous intention | 0.471 | 0.087 | 5.414 | 0.000 | Significant | Significant |
| H9 | Social Influence → Continuous intention | 0.624 | 0.758 | 0.824 | 0.410 | Not Significant | Not Significant |
| H10 | Facilitating conditions → Continuous intention | 0.325 | 0.338 | 0.873 | 0.382 | Not Significant | Not Significant |
| H12 | Attitude → Continuous intention | 0.175 | 0.076 | 2.288 | 0.022 | Significant | Not Significant |
Hypothetical analysis.
Regarding attitude formation, three of the four hypothesized relationships were supported. Effort expectancy (β = 0.620, p < 0.001) emerged as the strongest predictor of attitude, followed by performance expectancy (β = 0.521, p < 0.001) and hedonic motivation (β = 0.313, p < 0.001). However, habit did not significantly influence attitude (β = 0.138, p = 0.193), suggesting that automaticity alone does not shape students’ evaluative judgments of ML.
For continuous intention to use ML, the results revealed a more selective pattern. Habit was by far the strongest predictor (β = 0.445, p < 0.001), providing strong support for its dominance in driving sustained use. Hedonic motivation also demonstrated a substantial direct effect (β = 0.471, p < 0.001) on continuance intention, while attitude showed a significant but comparatively smaller influence (β = 0.175, p = 0.022).
Notably, several traditional UTAUT-2 constructs failed to reach statistical significance for continuous intention. Both performance expectancy (β = 0.028, p = 0.644) and effort expectancy (β = 0.063, p = 0.919) had non-significant direct effects, indicating that utilitarian considerations may have diminished in importance for continuance decisions. Similarly, social influence (β = 0.624, p = 0.410) and facilitating conditions (β = 0.325, p = 0.382) did not significantly affect students’ intention to continue using ML.
5 Discussion and implications
5.1 Summary of findings and theoretical contribution
This study successfully identified the key factors driving sustained mobile learning use among university students in the post-pandemic era. By applying the UTAUT-2 model, our findings reveal a significant restructuring of the factor landscape between mandatory adoption and voluntary continuance.
The most compelling finding is the dominant role of habit (β = 0.445, p < 0.001) in driving continuous intention, confirming the emerging global “habit-dominance” paradigm in post-crisis technology continuance. This result aligns with recent international studies (Ermilinda et al., 2024; Kelkay et al., 2025), suggesting that automated usage patterns formed during intensive pandemic use have become the primary driver of voluntary continuance.
However, the non-significant findings provide equally crucial insights into the evolving nature of technology acceptance. The lack of direct effects from performance expectancy (β = 0.028, p = 0.644) and effort expectancy (β = 0.063, p = 0.919) on continuous intention represents a fundamental departure from traditional Technology Acceptance Models. We posit that during the extended pandemic usage period, students accumulated sufficient firsthand experience with ML’s utility and ease of use, transforming these from conscious evaluations into background assumptions. Their influence appears to have been largely absorbed into habit formation, with habitual use now serving as the primary behavioral driver rather than ongoing utilitarian assessments.
Similarly, the non-significant effects of social influence (β = 0.624, p = 0.410) and facilitating conditions (β = 0.325, p = 0.382) reveal that ML use has become internalized and normalized in the post-pandemic context. The opinions of peers and basic technical support, while crucial during mandatory adoption, no longer serve as differentiating factors for continuance intention. This suggests ML has transitioned from a socially influenced behavior to a personal academic choice, with basic infrastructure perceived as an expected standard rather than a motivator.
The dual role of hedonic motivation is particularly noteworthy, as it demonstrates significant effects on both attitude (β = 0.313, p < 0.001) and continuous intention (β = 0.471, p < 0.001). This creates a crucial distinction: while utilitarian factors diminished, enjoyment persisted as a powerful direct driver. This underscores that sustained use requires not just automated behavior but also positive affective experiences.
This study contributes to theory by demonstrating how UTAUT-2 constructs evolve as technology use matures from adoption to embedded continuance. We provide a validated “Post-Adoption Continuance Model” where experiential factors (habit, hedonic motivation) supersede utilitarian factors (performance expectancy, effort expectancy) and external factors (social influence, facilitating conditions) in driving sustained use.
5.2 Practical implications
These findings translate into clear strategic directions for enhancing ML sustainability in the post-pandemic era:
For University Administrators and IT Departments:
Strategic Habit Formation: Move beyond promoting ML features to systematically designing for automatic use. Embed the platform into core academic workflows through consistent daily integration and predictable usage patterns.
Resource Reallocation: Given the non-significance of social influence, redirect efforts from peer persuasion campaigns toward initiatives that directly enhance user experience and habitual engagement.
Infrastructure as Baseline: Treat facilitating conditions as essential hygiene factors rather than key differentiators, ensuring reliable baseline support while focusing innovation on experiential aspects.
For Instructional Designers and Educators:
Dual-Pathway Design: Capitalize on hedonic motivation’s dual influence by incorporating engaging, interactive elements that simultaneously improve attitudes and drive continuance intention.
Routine Integration: Structure course interactions to create consistent usage patterns that gradually build habitual engagement, reducing reliance on conscious decision-making.
Experience over Utility: Emphasize enjoyable learning experiences rather than focusing primarily on efficiency gains when promoting ML tools to students.
For ML Platform Developers:
Habit-Centric Architecture: Implement features that encourage routine engagement, such as personalized notifications, streamlined workflows, and progress tracking that reinforces regular use.
Affective Experience Design: Prioritize enjoyable user interactions and interface design, recognizing that hedonic motivation is a primary sustainer of long-term use alongside habit.
Seamless Integration: Focus on reducing friction points to support the automaticity of use, recognizing that ease of use has become an expectation rather than a motivator.
6 Conclusion
This study successfully addressed the research gap concerning the drivers of continuous ML intention in the post-pandemic era within Saudi higher education. The findings provide a nuanced understanding of how technology acceptance factors evolve after periods of intensive use.
The results demonstrate that habit has emerged as the cornerstone of post-pandemic ML continuance, confirming the global shift toward automated behavioral patterns as the primary driver of sustained technology use. The strong influence of hedonic motivation on both attitude and continuance intention highlights the enduring importance of enjoyment in the learning technology experience.
Equally significant are the factors that were not significant for continuance intention. The lack of direct effects from performance expectancy and effort expectancy signals a fundamental transition in what drives sustained use compared to initial adoption. The insignificance of social influence and facilitating conditions further suggests that ML has transitioned from an externally influenced behavior to an individually determined practice.
These findings collectively indicate that post-pandemic ML continuance is driven primarily by automated patterns formed through past use, complemented by intrinsic enjoyment, rather than conscious utility evaluations or external pressures.
For educational institutions, this study offers a clear directive: sustainable ML integration requires strategies that deliberately foster habitual use and enjoyable experiences, moving beyond the utility-focused approaches that sufficed during emergency adoption. The future of mobile learning depends not on compelling features alone, but on embedded practices and positive experiences that students willingly maintain.
For future research, longitudinal studies tracking the evolution of these factors from adoption to long-term continuance would be valuable. Investigating the specific instructional design features that most effectively foster habit formation could yield refined strategies for sustainable technology integration in education.
7 Limitations and future studies
While this study provides valuable insights, several limitations should be acknowledged, which also present avenues for future research.
(1) Sampling and Generalizability: The study employed a purposive sampling strategy from a single college (Computer Science and Information Technology) at one university in Saudi Arabia. This specific sample characteristic limits the generalizability of the findings. The results are most representative of students within a similar technological and cultural context. The perspectives of students from humanities, health sciences, or other disciplines, as well as those from other regions or countries, may differ. Future studies should employ stratified random sampling across multiple universities and diverse academic disciplines to enhance the external validity and generalizability of the findings.
(2) Methodological Scope and Cross-Sectional Design: This research utilized a cross-sectional survey, capturing data at a single point in time. Consequently, it can demonstrate relationships between factors but cannot definitively establish causality. Furthermore, the exclusive reliance on a quantitative approach, while effective for testing the proposed model, limits the depth of understanding regarding the underlying reasons for students’ continuance intention. As suggested, future research would benefit from a mixed-methods approach, integrating qualitative interviews or focus groups to provide rich, contextual explanations for the statistical relationships uncovered here.
(3) Source of Data and Potential Biases: The study relied exclusively on self-reported data, which is susceptible to biases such as social desirability bias (where respondents answer in a way they believe is socially acceptable) and common method variance. Using a single questionnaire across all constructs may have artificially inflated the relationships among them. Future studies could mitigate this by collecting data from different sources (e.g., pairing student surveys with actual platform usage metrics) or by temporally separating the measurement of predictor and outcome variables.
(4) Unexplored Constructs and Perspectives: The study focused on a student-centric view based on the UTAUT-2 framework. The perspective of teachers, who are crucial actors in the educational ecosystem, was not investigated. Their acceptance, habits, and facilitation skills are likely critical to the successful and sustained integration of ML. Future studies should develop a complementary model to explore the drivers of continuance intention among educators. Additionally, while other models like TAM were considered, the comprehensive nature of UTAUT-2 was deemed most suitable; however, integrating constructs from other theories could provide a more holistic understanding.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by Research Ethics Committee at King Faisal University (Ref no: KFU-REC-2024-JUL-ETHICS2446). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
AM: Project administration, Formal analysis, Writing – original draft, Software, Data curation, Methodology, Visualization, Investigation, Validation, Funding acquisition, Supervision, Writing – review & editing, Resources, Conceptualization.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This study was supported by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia (Project No.: KFU253859).
Conflict of interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author declares that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
AfzalJ.AnwarG. (2023). An empirical study on academic sustainability of mobile learning at university level. Pak. J. Educ. Res.6, 78–91. doi: 10.52337/pjer.v6i2.788
2
AgustyaningrumN.SariR. N.AbadiA. M.MahmudiA. (2021). Dominant factors that cause students’ difficulties in learning abstract algebra: a case study at a university in Indonesia. Int. J. Instr.14, 847–866. doi: 10.29333/iji.2021.14151a
3
AhmadS. F.AlamM. M.RahmatM. K.MubarikM. S.HyderS. I. (2022). Academic and administrative role of artificial intelligence in education. Sustainability14:1101. doi: 10.3390/su14031101
4
Al MulhemA. (2020a). Investigating the effects of quality factors and organizational factors on university students’ satisfaction of e-learning system quality. Cogent Educ.7:1787004. doi: 10.1080/2331186x.2020.1787004
5
Al MulhemA. (2020b). Exploring the key factors in the use of an e-learning system among students at King Faisal University, Saudi Arabia. Int. J. Interact. Mob. Technol.14, 19–37. doi: 10.3991/ijim.v14i03.11576
6
Al MulhemA.AlmaiahM. (2021). A conceptual model to investigate the role of mobile game applications in education during the COVID-19 pandemic. Electronics10:2106. doi: 10.3390/electronics10172106
7
AlamriM. M.AlmaiahM. A.Al-RahmiW. M. (2020a). Social media applications affecting students’ academic performance: a model developed for sustainability in higher education. Sustainability12:6471. doi: 10.3390/su12166471
8
AlamriM. M.AlmaiahM. A.Al-RahmiW. M. (2020b). The role of compatibility and task-technology fit (TTF): on social networking applications (SNAs) usage as sustainability in higher education. IEEE Access8, 161668–161681. doi: 10.1109/access.2020.3021944
9
AlatniB. S.AbubakarI. R.IqbalS. A. (2021). COVID-19 and rapid course adaptations in Saudi Arabia: an experiential learning and recommendations for online education. Front. Psychol.12:643203. doi: 10.3389/fpsyg.2021.643203
10
Al-EmranM.ArpaciI.SalloumS. A. (2020). An empirical examination of continuous intention to use m-learning: an integrated model. Educ. Inf. Technol.25, 2899–2918. doi: 10.1007/s10639-019-10094-2
11
AlkabaaA. S. (2022). Effectiveness of using E-learning systems during COVID-19 in Saudi Arabia: experiences and perceptions analysis of engineering students. Educ. Inf. Technol.27, 10625–10645. doi: 10.1007/s10639-022-11054-z
12
AlksasbehM.AbuhelalehM.AlmaiahM. A.Al-JaafrehM.Abu karakaA. (2019). Towards a model of quality features for mobile social networks apps in learning environments: an extended information system success model. Int. J. Interact. Mob. Technol.13, 75–96. doi: 10.3991/ijim.v13i05.9791
13
AlmaiahM. A.Abdul JalilM. (2014). Investigating students’ perceptions on mobile learning services. Int. J. Interact. Mob. Technol.8, 31–36. doi: 10.3991/ijim.v8i4.3965
14
AlmaiahM. A.Al MulhemA. (2020). Thematic analysis for classifying the main challenges and factors influencing the successful implementation of e-learning system using NVivo. Int. J. Adv. Trends Comput. Sci. Engineer.9, 32–44. doi: 10.30534/ijatcse/2020/22912020
15
AlmaiahM. A.AlfaisalR.SalloumS. A.Al-OtaibiS.Al SawafiO.Al-MaroofR. S.et al. (2022d). Determinants influencing the continuous intention to use digital technologies in higher education. Electronics11:2827. doi: 10.3390/electronics11182827
16
AlmaiahM. A.AlfaisalR.SalloumS. A.Al-OtaibiS.ShishaklyR.LutfiA.et al. (2022e). Integrating teachers’ TPACK levels and students’ learning motivation, technology innovativeness, and optimism in an IoT acceptance model. Electronics11:3197. doi: 10.3390/electronics11193197
17
AlmaiahM. A.Al-KhasawnehA.AlthunibatA. (2020). Exploring the critical challenges and factors influencing the E-learning system usage during COVID-19 pandemic. Educ. Inf. Technol.25, 5261–5280. doi: 10.1007/s10639-020-10219-y
18
AlmaiahM. A.Al-KhasawnehA.AlthunibatA.AlmomaniO. (2021a). “Exploring the main determinants of mobile learning application usage during Covid-19 pandemic in Jordanian universities” in Emerging technologies during the era of COVID-19 pandemic. eds. HassanienA. E.HaqiqA.RodriguesP. J. S. G.TolbaA. F.ElhosenyM. (Cham, Switzerland: Springer), 275–290.
19
AlmaiahM. A.AlmomaniO.Al-KhasawnehA.AlthunibatA. (2021b). “Predicting the acceptance of mobile learning applications during COVID-19 using machine learning prediction algorithms” in Emerging technologies during the era of COVID-19 pandemic. eds. HassanienA. E.HaqiqA.RodriguesP. J. S. G.TolbaA. F.ElhosenyM. (Cham, Switzerland: Springer), 319–332.
20
AlmaiahM. A.Al-loziE.Al-KhasawnehA.ShishaklyR.NachoukiM. (2021c). Factors affecting students’ acceptance of mobile learning application in higher education during COVID-19 using ANN-SEM modelling technique. Electronics10:3121. doi: 10.3390/electronics10243121
21
AlmaiahM. A.HajjejF.LutfiA.Al-KhasawnehA.AlkhdourT.AlmomaniO.et al. (2022a). A conceptual framework for determining quality requirements for mobile learning applications using Delphi method. Electronics11:788. doi: 10.3390/electronics11050788
22
AlmaiahM. A.Al-OtaibiS.LutfiA.AlmomaniO.AwajanA.AlsaaidahA.et al. (2022b). Employing the TAM model to investigate the readiness of M-learning system usage using SEM technique. Electronics11:1259. doi: 10.3390/electronics11081259
23
AlmaiahM. A.JalilM. A.ManM. (2016). Extending the TAM to examine the effects of quality features on mobile learning acceptance. J. Comput. Educ.3, 453–485. doi: 10.1007/s40692-016-0074-1
24
AlmaiahM. A.NasereddinY. (2020). Factors influencing the adoption of e-government services among Jordanian citizens. Electron. Govern. Int. J.16, 1–19. doi: 10.1504/eg.2020.10019265
25
Al-MaroofR. S.AlnazzawiN.AkourI. A.AyoubiK.AlhumaidK.AlAhbabiN. M.et al. (2021a). The effectiveness of online platforms after the pandemic: will face-to-face classes affect students’ perception of their behavioural intention (BIU) to use online platforms?Informatics8:83. doi: 10.3390/informatics8040083
26
Al-MaroofR. S.AyoubiK.AlhumaidK.AburayyaA.AlshuridehM.AlfaisalR.et al. (2021b). The acceptance of social media video for knowledge acquisition, sharing and application: a comparative study among YouTube users and TikTok users’ for medical purposes. Int. J. Data Netw. Sci.5, 197–214. doi: 10.5267/j.ijdns.2021.6.013
27
Al-MaroofR. S.SalloumS. A. (2020). “An integrated model of continuous intention to use of Google classroom” in Recent advances in intelligent systems and smart applications. ed. SalloumS. A. (Cham, Switzerland: Springer), 311–335.
28
AlqahtaniM. H. (2022). Every cloud has a silver lining: the experience of online learning in English language classes at Saudi universities in the post–COVID-19 era. Front. Educ.7:1027121. doi: 10.3389/feduc.2022.1027121
29
AlshathryS.AlojailM. (2024). Evaluating post-pandemic undergraduate student satisfaction with online learning in Saudi Arabia: the significance of self-directed learning. Appl. Sci.14:8889. doi: 10.3390/app14198889
30
AlsheibaniS.CheungY.MessomC.. (2019). Factors inhibiting the adoption of artificial intelligence at organizational-level: A preliminary investigation. Paper presented at the Americas Conference on Information Systems, Cancun, Mexico. Available online at: https://aisel.aisnet.org/amcis2019/adoption_diffusion_IT/adoption_diffusion_IT/2/ (Accessed May 30, 2025).
31
AlsyoufA.LutfiA.Al-BsheishM.JarrarM.Al-MugheedK.AlmaiahM. A.et al. (2022). Exposure detection applications acceptance: the case of COVID-19. Int. J. Environ. Res. Public Health19:7307. doi: 10.3390/ijerph19127307
32
AlthunibatA.AlmaiahM. A.AltarawnehF. (2021). Examining the factors influencing the mobile learning applications usage in higher education during the COVID-19 pandemic. Electronics10:2676. doi: 10.3390/electronics10212676
33
AlturkiU.AldraiweeshA. (2022). Students’ perceptions of the actual use of mobile learning during COVID-19 pandemic in higher education. Sustainability14:1125. doi: 10.3390/su14031125
34
ArpaciI.BasolG. (2020). The impact of preservice teachers’ cognitive and technological perceptions on their continuous intention to use flipped classroom. Educ. Inf. Technol.25, 3503–3514. doi: 10.1007/s10639-020-10104-8
35
BabatundeS. A.AjapeM. K.IsaK. D.KuyeO.OmolehinwaE. O.MuritalaS. A. (2021). Ease of doing business index: an analysis of investors practical view. J. Econ.17, 101–123. doi: 10.21831/economia.v17i1.33941
36
BachmairB.PachlerN. (2015). Sustainability for innovative education – the case of mobile learning. J. Interact. Media Educ.2015, 1–12. doi: 10.5334/jime.ay
37
ChaudhryM.KazimE. (2021). Artificial intelligence in education (Aied) a high-level academic and industry note 2021. SSRN Electron. J.2:157–165. doi: 10.2139/ssrn.3833583
38
Chávez HertingD.Cladellas ProsR.Castelló TarridaA. (2020). Habit and social influence as determinants of PowerPoint use in higher education: a study from a technology acceptance approach. Interact. Learn. Environ.31, 497–513. doi: 10.1080/10494820.2020.1799021
39
ChenX.ZouD.XieH.ChengG.LiuC. (2022). Two decades of artificial intelligence in education: contributors, collaborations, research topics, challenges, and future directions. J. Educ. Technol. Soc.25, 28–47.
40
García de Blanes SebastiánM.Sarmiento GuedeJ. R.Azuara GrandeA.FilipeA. F. (2024). UTAUT-2 predictors and satisfaction: implications for mobile-learning adoption among university students. Educ. Inf. Technol.30, 3201–3237. doi: 10.1007/s10639-024-12927-1
41
DerbaliA.LtaifaM. B. (2022). “The impact of E-learning on education outcomes in Saudi universities considering the COVID-19 epidemic” in Technologies, artificial intelligence and the future of learning post-COVID-19. eds. HamdanA.HassanienA. E.MesconT.AlareeniB. (Cham, Switzerland: Springer), 369–392.
42
ElareshiM.HabesM.YoussefE.SalloumS. A.AlfaisalR.ZianiA. (2022). SEM-ANN-based approach to understanding students’ academic-performance adoption of YouTube for learning during Covid. Heliyon8:e09236. doi: 10.1016/j.heliyon.2022.e09236
43
ElumalaiK. V.SankarJ. P.RK.JohnJ. A.MenonN.AlqahtaniM. S. M.et al. (2020). Factors affecting the quality of E-learning during the COVID-19 pandemic from the perspective of higher education students. J. Inf. Technol. Educ. Res.19, 731–753. doi: 10.28945/4628
44
ErmilindaL.HandarkhoY. D.EmanuelA. W. R. (2024). Factors influencing student intention to continue using e-learning platform post Covid-19 pandemic: case study of University of Nusa Nipa Indonesia. Procedia Comput. Sci.234, 1043–1052. doi: 10.1016/j.procs.2024.03.098
45
FornellC.LarckerD. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39–50. doi: 10.1177/002224378101800104
46
HairJ. F.HollingsworthC. L.RandolphA. B.ChongA. Y. L. (2017). An updated and expanded assessment of PLS-SEM in information systems research. Ind. Manag. Data Syst.117, 442–458. doi: 10.1108/imds-04-2016-0130
47
HairJ. F.HultG. T. M.RingleC. M.SarstedtM. (2022). A primer on partial least squares structural equation modeling (PLS-SEM). 3rd Edn. Thousand Oaks: Sage.
48
HassanA. (2022). “The experience of the Kingdom of Saudi Arabia in the field of E-learning during the coronavirus pandemic” in Technologies, artificial intelligence and the future of learning post-COVID-19. eds. HamdanA.HassanienA. E.MesconT.AlareeniB. (Cham, Switzerland: Springer), 355–368.
49
HoJ. C. (2021). Disruptive innovation from the perspective of innovation diffusion theory. Tech. Anal. Strat. Manag.34, 363–376. doi: 10.1080/09537325.2021.1901873
50
HooksD.DavisZ.AgrawalV.LiZ. (2021). Exploring factors influencing technology adoption rate at the macro level: a predictive model. SSRN Electron. J., 68, 101826. doi: 10.2139/ssrn.3943624
51
KelkayJ. M.WubanteS. M.AntenehD. S.TakiloM. K.GebeyehuC. D.AlamerawT. A.et al. (2025). Intention to use elearning-based continuing professional development and its predictors among healthcare professionals in Amhara region referral hospitals, Ethiopia, 2023: using modified UTAUT-2 model. BMC Health Services Research, 25, 178. doi: 10.1186/s12913-025-12317-4
52
KosibaJ. P. B.OdoomR.BoatengH.TwumK. K.Abdul-HamidI. K. (2022). Examining students’ satisfaction with online learning during the Covid-19 pandemic – an extended UTAUT2 approach. J. Furth. High. Educ.46, 988–1005. doi: 10.1080/0309877x.2022.2030687
53
LutfiA. (2022). Factors influencing the continuance intention to use accounting information system in Jordanian SMEs from the perspectives of UTAUT: top management support and self-efficacy as predictor factors. Economies10:75. doi: 10.3390/economies10040075
54
LutfiA.Al-KhasawnehA. L.AlmaiahM. A.AlsyoufA.AlrawadM. (2022a). Business sustainability of small and medium enterprises during the COVID-19 pandemic: the role of AIS implementation. Sustainability14:5362. doi: 10.3390/su14095362
55
LutfiA.SaadM.AlmaiahM. A.AlsaadA.Al-KhasawnehA.AlrawadM.et al. (2022b). Actual use of mobile learning technologies during social distancing circumstances: case study of King Faisal University students. Sustainability14:7323. doi: 10.3390/su14127323
56
MalikG.TayalD. K.VijS. (2019). “An analysis of the role of artificial intelligence in education and teaching” in Recent findings in intelligent computing techniques. eds. SaP. K.BakshiS.HatzilygeroudisI.SahooM. N. (Singapore: Springer), 407–417. doi: 10.1007/978-981-10-8639-7_42
57
MartinsM.FariasJ.AlbuquerqueP. H.PereiraD. (2018). Adoption of technology for reading purposes: a study of E-books acceptance. Braz. Bus. Rev.15, 568–588. doi: 10.15728/bbr.2018.15.6.4
58
NezamdoustS.AbdekhodaM.RahmaniA. (2022). Determinant factors in adopting mobile health application in healthcare by nurses. BMC Med. Inform. Decis. Mak.22, 47–11. doi: 10.1186/s12911-022-01784-y
59
NtsifulA.KwartengM. A.PilíkM.OsakweC. N. (2022). Transitioning to online teaching during the pandemic period: the role of innovation and psychological characteristics. Innov. High. Educ.48, 197–218. doi: 10.1007/s10755-022-09613-w
60
NunnallyJ. C. (1978). Psychometric theory (2nd ed.). McGraw-Hill.
61
RingleC. M.WendeS.BeckerJ. M. (2015). SmartPLS 3. SmartPLS GmbH. Available online at: http://www.smartpls.com (Accessed April 12, 2025).
62
RohanR.PalD.FunilkulS.ChutimaskulW.EamsinvattanaW. (2021). How gamification leads to continued usage of MOOCs? A theoretical perspective. IEEE Access9, 108144–108161. doi: 10.1109/access.2021.3102293
63
RudhumbuN. (2022). Applying the UTAUT2 to predict the acceptance of blended learning by university students. Asian Assoc. Open Univ. J.17, 15–36. doi: 10.1108/aaouj-08-2021-0084
64
SalloumS. A.AlhamadA. Q. M.Al-EmranM.MonemA. A.ShaalanK. (2019). Exploring students’ acceptance of e-learning through the development of a comprehensive technology acceptance model. IEEE Access7, 128445–128462. doi: 10.1109/access.2019.2939467
65
SalloumS. A.ShaalanK. (2018). “Adoption of E-book for university students” in Proceedings of the international conference on advanced intelligent systems and informatics 2018. eds. HassanienA. E.TolbaM. F.ShaalanK.AzarA. T. (Cham, Switzerland: Springer), 481–494. doi: 10.1007/978-3-319-99010-1_44
66
SanduN.GideE.. (2019). Adoption of AI-Chatbots to enhance student learning experience in higher education in India. 2019 18th International Conference on Information Technology Based Higher Education and Training (ITHET), Magdeburg, Germany.
67
SetirekA. C.TanrikuluZ. (2015). Significant developmental factors that can affect the sustainability of mobile learning. Procedia. Soc. Behav. Sci.191, 2089–2096. doi: 10.1016/j.sbspro.2015.04.336
68
SmedaA.ShiratuddinM. F.WongK. W. (2018). A structural equation modelling approach for adoption of e-book amongst mathematics and statistics (MAS) students at higher education institutions in Libya. Int. J. Inf. Learn. Technol.35, 240–254. doi: 10.1108/ijilt-05-2017-0043
69
SteynA. A.van SlykeC.DickG.TwinomurinziH.AmusaL. B. (2024). Student intentions to continue with distance learning post-COVID: an empirical analysis. PLoS One19:e0293065. doi: 10.1371/journal.pone.0293065
70
TaamnehA.AlsaadA.ElrehailH.Al-OkailyM.LutfiA.SergioR. P. (2022). University lecturers acceptance of moodle platform in the context of the COVID-19 pandemic. Glob. Know. Memory Commun.72, 666–684. doi: 10.1108/gkmc-05-2021-0087
71
TahatK. M.Al-SarayrahW.SalloumS. A.HabesM.AliS. (2021). “The influence of YouTube videos on the learning experience of disabled people during the COVID-19 outbreak” in Advances in data science and intelligent data communication technologies for COVID-19. eds. HassanienA. E.HaqiqA. (Cham, Switzerland: Springer), 239–252. doi: 10.1007/978-3-030-77302-1_13
72
UNESCO. (2015). Incheon declaration: education 2030: towards inclusive and equitable quality education and lifelong learning for all. UNESCO. Available online at: https://unesdoc.unesco.org/ark:/48223/pf0000245656 (Accessed March 10, 2025).
73
UrbachN.AhlemannF. (2010). Structural equation modeling in information systems research using partial least squares. J. Inf. Technol. Theory Appl.11, 5–40.
74
VenkateshV.MorrisM. G.DavisG. B.DavisF. D. (2003). User acceptance of information technology: toward a unified view. MIS Q.27, 425–478. doi: 10.2307/30036540
75
VenkateshV.ThongJ. Y. L.XuX. (2012). Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology. MIS Q.36, 157–178. doi: 10.2307/41410412
76
WilsonC.van SteenT.AkinyodeC.BrodieZ.ScottG. G. (2021). To sext or not to sext. The role of social-cognitive processes in the decision to engage in sexting. J. Soc. Pers. Relat.38, 1410–1429. doi: 10.1177/0265407521995884
77
YangW. (2022). Artificial intelligence education for young children: why, what, and how in curriculum design and implementation. Comput. Educ.3:100061. doi: 10.1016/j.caeai.2022.100061
78
ZacharisG.NikolopoulouK. (2022). Factors predicting university students’ behavioral intention to use eLearning platforms in the post-pandemic normal: an UTAUT2 approach with ‘learning value. Educ. Inf. Technol.27, 12065–12082. doi: 10.1007/s10639-022-11116-2
79
ZhengH.HanF.HuangY.WuY.WuX. (2025). Factors influencing behavioral intention to use e-learning in higher education during the COVID-19 pandemic: a meta-analytic review based on the UTAUT2 model. Educ. Inf. Technol.30, 12015–12053. doi: 10.1007/s10639-024-13299-2
Summary
Keywords
mobile learning, sustainability, UTAUT-2, e-learning, university students, post-COVID-19 time
Citation
Al Mulhem AA (2025) A model for sustainable mobile education beyond the COVID-19 pandemic. Front. Educ. 10:1657635. doi: 10.3389/feduc.2025.1657635
Received
01 July 2025
Accepted
28 October 2025
Published
25 November 2025
Volume
10 - 2025
Edited by
Joseline Santos, Bulacan State University, Philippines
Reviewed by
Wanda Nugroho Yanuarto, Muhammadiyah University of Purwokerto, Indonesia
Noor Saadiah Mohd Ali, International Islamic University Malaysia, Malaysia
Updates
Copyright
© 2025 Al Mulhem.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ahmed Abdulhameed Al Mulhem, ahamulhem@kfu.edu.sa
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.