Modelling the effects of emotional engagement and peer interaction on the continuous intention to use asynchronous e-learning

Theoretical framework

This study uses TAM as a theorical framework. Technology acceptance means the extent which users intend to utilise and keep using any new technology. TAM is a validated model which has been used to analyse instructors and students’ behavioural intention to use new technologies in several different educational contexts, such as teachers’ intention with using games (Yeo, Rutherford & Campbell, 2022), LMS (Alshammari, 2020) and mobile-based assistive technology in learning languages (Hsu & Lin, 2022).

The TAM theoretical framework contains four main factors: perceived usefulness (PU), ease of use (PEU), attitude (ATT) and the behaviour intention to use a technology (BI). PEU means the extent which users feel that using particular technology will be easy and require little effort (Davis, 1989). PU means the extent which users believe that using specific technology could be beneficial and could enhance their tasks and performance (Davis, 1989). Furthermore, ATT refers to a user’s negative or positive feelings regarding using a technology. It was recommended that the attitude construct should be eliminated from the TAM due to the direct effects of PU and PEU on users’ intention (Davis, 1989). BI is defined as the intention of users toward using particular technology or application (Davis, 1989).

Emotional engagement with TAM

Emotional engagement refers to students’ emotional interaction with their learning in either a positive sense (such as when learning aligns with students’ identities, satisfaction, interests and values) or a negative sense (for instance, when learning leads to anxiety) (Chan, Kam & Wong, 2022). Several previous studies revealed that, emotional engagement showed to be as indicator for students’ high academic achievement. Kuo et al. (2014) determined that increasing students’ emotional engagement led to improved learning outcomes. Similarly, Kahu & Nelson (2018) established that emotional engagement could predict the academic success of students. Furthermore, Wang et al. (2015) assessed the relationships between emotional engagement in LMSs and students’ academic performance and found that emotional engagement with learning activities through LMSs had a direct influence on students’ academic and learning performance. However, emotional engagement’s effect on students’ continuous intention for using asynchronous e-learning remains unknown. Based on this analysis of the literature, hypothesis is formulated as following:

H1: Emotional engagement affects students’ perceptions of the usefulness of asynchronous e-learning.

H2: Emotional engagement affects students’ perceptions of the ease of use of asynchronous e-learning.

Peer interaction with TAM

Peer interaction means the interactions between learners with their instructors or with other learners. When students interact with other students and instructors, they can improve their ability to build knowledge (Liaw, Huang & Chen, 2007). Liu et al. (2010) examined the constructs which may influence students’ utilises of e-learning communities. They found that the peer interaction of learners and the perceived ease of online learning community had affected students’ use of that LMS. Similarly, Rose et al. (2015) determined that the peer interaction of learners affected students’ intention to utilise LMS. Thus, the hypothesis is formulated as following:

H3: Peer interaction affects students’ perceptions of the usefulness of asynchronous e-learning.

H4: Peer interaction affects students’ perceptions of the ease of use of asynchronous e-learning.

TAM constructs

Regarding TAM constructs, previous studies have shown that students’ perceived ease of using particular technology directly affect their perceptions of the usefulness of that technology. Furthermore, PU and PEU affect the intention of using a technology. These results have been confirmed in several contexts, such as in language learning using mobile phones (Hsu & Lin, 2022), LMS (Alshammari, 2018) and LMSs (Yalcin & Kutlu, 2019). Accordingly, hypothesis is formulated as following:

H5: Perceived ease of use affects students’ perceptions of the usefulness of asynchronous e-learning.

H6: Perceived ease of use affects students’ continuous intention to use asynchronous e-learning.

H7: Perceived usefulness affects students’ continuous intention to use asynchronous e-learning.

Figure 1 shows the proposed mode.

Methodology

Research design

A quantitative design approach is used in this study and data was collected by questionnaires. Quantitative research aims to explain different phenomena by examining numerical data that is analysed using a mathematical technique, specifically, a statistical approach (Creswell, 2013). The quantitative research design is systematic, formal and objective and aims to assess effects, causes and relationships between variables (Queirós, Faria & Almeida, 2017). Thus, design of quantitative research is suitable for this research, since the study aim at proposing a theoretically extended TAM and exploring the effects of external factors (namely, emotional engagement and peer interaction) and TAM constructs on students’ continuous intention to use asynchronous e-learning.

Demographic information analysis

A total of 259 students who had experience in using asynchronous e-learning participated in the survey. Respondents’ demographic information is shown in Table 1. Regarding the respondents’ gender, 145 (56.0%) were male students, and 114 (44.0%) were female students. In relation to the respondents’ academic programme, the vast majority of the students (215; 83.0%) were enrolled in a bachelor’s programme. Just 24 (9.3%) and 20 (7.7%) students were enrolled in diploma and master’s programmes, respectively. Most of the participants were studying at the College of Business Administration (94; 36.3%), followed by the College of Computer Science and Engineering (49; 18.9%) and then the College of Arts (45; 17.4%). A small number of students (14; 5.4%) were studying at the Applied College.

Confirmatory factor analysis

Pooled CFA is the most convenient technique used to develop and validate a measurement model, as it has the power to account for all correlations of constructs and handle measurement errors (Hair et al., 2010; Awang, 2015). During CFA, validities of construct, convergent and discriminant should be evaluated to develop measurements (Awang, 2015). Figure 2 presents CFA output.

Construct validity is met once the model fitness reaches the threshold values suggested in the literature (Awang, 2015). The values of the model fitness indices met the recommended levels, as Table 2 presented.

Convergent validity is achieved once composite reliability (CR) score is 0.6 or greater and AVE scores 0.5 or greater (Awang, 2015). The CR and AVE values displayed in Table 3 show that the suggested values for convergent validity were met.

Discriminant validity means that each construct is different than other related constructs approved by statistical and empirical standards (Hair Jr et al., 2021). For assessing discriminant validity of constructs, the AVE square root should be assessed to compare their correlations in the model (Fornell & Larcker, 1981). The bold values in Table 4 show the square root of AVE. All bold values greater than other values in their columns or rows. Therefore, discriminant validity was achieved (Awang, 2015).

Standardised estimate

The SEM presents two kinds of outputs. The standardised estimate is essential for assessing the beta coefficients of variables, factor loading and the R square of dependent variables (Awang, 2015). The unstandardised estimate is applied for calculating critical ratio for testing hypotheses. The standardised estimate results are presented in Fig. 3.

The R square of dependent variable–continuous intention to utilise asynchronous e-learning–was 0.88, which confirms that 88% of continuous intention of using asynchronous e-learning could be explained by other variables. This result confirms the high explanatory power in the proposed model for explaining the phenomenon of students accepting and using asynchronous e-learning and the factors that affect this. Cohen (1988) mentioned that, the value of R square less than 0.12 indicates low explanatory power, whereas ranges in the middle of 0.13 and 0.25 refers to medium power and values higher than 0.25 indicate high explanatory power. Thus, this study produced a model with high explanatory power that is able to clearly indicate which factors contribute to continuous intention to utilise asynchronous e-learning.

Unstandardised estimate

The regression weight which used for assessing critical ratio and testing hypotheses is calculated by applying an unstandardised estimate. Figure 4 presents unstandardised estimate results.

Results of hypothesis testing

The results show that emotional engagement was insignificant in affecting perceived usefulness (β = 0.156, P > 0.05); therefore, H1 is rejected. However, emotional engagement was significant in affecting perceived ease of use (β = 0.905, P < 0.001). Accordingly, H2 is supported. Furthermore, peer interaction has a significant effect on both PU and PEU (β = 0.206, P < 0.05; β = −0.149, P < 0.05), which supports H3 and H4. Moreover, PEU was significant in affecting PU (β = 0.611, P <0.05), Therefore, supporting H5. Finally, both PEU and PU significantly affected students’ continuous intention to use asynchronous e-learning (β = 0.823, P < 0.05; β = −0.187, P < 0.05). Thus, H6 and H7 are confirmed. Table 5 presents hypothesis testing results.

Furthermore, results confirmed the indirect effect of emotional engagement on students’ continuous intention to use asynchronous e-learning (β = 0.878, P < 0.05), the indirect effect value is shown in Table 6.

Implications

This study has several theoretical implications. Few studies have examined the continuous intention to use asynchronous e-learning. Lu, Pang & Shadiev (2023) argued for extending TAM and examining the effects of emotional engagement and peer interaction on students’ continuous intention to use asynchronous e-learning. It was not yet clear whether these two factors affect students’ asynchronous e-learning. This study filled that gap and contributed to the literature by proposing a theoretical model incorporating emotional engagement and peer interaction and examining their effects on students’ asynchronous e-learning. To the best of our knowledge, this is the first study that proposes a theoretical model incorporating emotional engagement and peer interaction along with TAM constructs to examine their effects on students’ continuous intention to use asynchronous e-learning.

Moreover, the study findings have practical implications. First, it is important to ensure the easiness and usefulness of asynchronous e-learning, as these factors play a big role in students’ continuous intention to use asynchronous e-learning. That can be achieved by providing multiple resources for learning, useful content and different technologies. Furthermore, training students on how to navigate and use e-learning materials will make using asynchronous e-learning easy, which will lead to students perceiving it as useful. Second, there is a need to enhance the positive emotional engagement of students in asynchronous e-learning in terms of their interests, values and happiness. Third, instructors should pay attention to students’ peer interactions with their instructors and fellow students during asynchronous e-learning and provide collaborative learning activities that require interactions among students, such as discussions on discussion forums or blogs. This will increase students’ peer interactions and their continuous intention to use asynchronous e-learning.

Limitations

Some limitations of this study can be identified. To start with, the study data were self-reported by students and were subjective. Other forms of data collection, such as observations, need to be taking in consideration for future studies. Furthermore, participants were enrolled in one university in Saudi Arabia. Future studies should include participants from different universities in Saudi Arabia or in other countries for comparing the findings with this study and generalising findings. Finally, this study incorporated two key external factors within TAM: emotional engagement and peer interaction. Future studies should incorporate other important factors, such as the instructional design and accessibility of asynchronous e-learning.