As has been noted, the present study builds on the conservation of resource theory (Hobfoll, 1989) to understand the link between self-regulation and behavioral intention to reuse. Conservation of resource theory assumes that individuals tend to protect and retain the resources they value. The resource could be anything that people think is valuable (Hobfoll, 1989). It suggests that when individuals lack resources, they tend to withdraw behavioral intention to minimize their loss. By contrast, individuals are interested in investing in more resources when such an investment generates returns. Because behavioral intention to reuse is considered a type of behavioral attitude toward acceptance and satisfaction (Chiu et al., 2005), which reflects a subjective expectation of how many resources learners can obtain by using e-learning systems, we hypothesized that self-regulation is positively related to behavioral intention to reuse.

Behavioral intention refers to the degree to which an individual intends to perform some specific future behavior (Davis and Warshaw, 1992). In the context of e-learning, behavioral intention to reuse has been defined as the likelihood that learners will use an e-learning system again when it is made available to them (Budu et al., 2018). As learners’ willingness to adopt an e-learning system determines the success of any e-learning system (Li et al., 2012; Almaiah, 2018), understanding the factors that influence behavioral intention to reuse has been the focus of numerous e-learning studies (Al-Emran and Teo, 2020; Zhang et al., 2020). Previous research has identified several variables influencing learners’ decisions to use e-learning systems. These include perceived usefulness and perceived ease of use (Liaw, 2008); self-efficacy and attitudes toward use (Amin et al., 2015); the experience of e-learning (Mailizar et al., 2021); and social influence (Liaw, 2008).

Self-regulation is an important factor in effective learning and is widely used to explain learning outcomes in e-learning research. Self-regulated behaviors include but are not limited to environment structuring, goal setting, time management, help-seeking, task strategies, and self-evaluation (Barnard et al., 2009). According to the conservation of resource theory, the more learners perceive e-learning systems to be beneficial, the more they will use them. Because self-regulation enables learners to identify e-learning resources to achieve their goals (Wong et al., 2019), it can be assumed that it will increase their desire and belief to reuse them. Broadbent and Poon (2015) carried out a comprehensive analysis of self-regulation and concluded that self-regulation can help learners to achieve academic success in higher education settings. Various self-regulation outcomes have also been examined in previous studies (Panadero, 2017; van Alten et al., 2020a), for instance, satisfaction (van Alten et al., 2020b), learning performance (Zheng and Zhang, 2020), and academic achievement (Broadbent and Poon, 2015). Based on the above literature, we proposed the following hypothesis:

The concept of study engagement has received increased attention in the e-learning literature (Rodgers, 2008; Moubayed et al., 2020). Study engagement is a psychological state of vigor, absorption, and dedication toward learning tasks (Lamborn et al., 1992) and reflects “the amount of effort dedicated to educational activities that bring out ideal performance” (Hu and Kuh, 2002, p. 555). Individual resources are an essential component of an individual’s engagement in their studies (Ryan, 2005). Previous literature has identified several resources that facilitate student engagement, such as interest in learning, belongingness (Hu and Kuh, 2002), deep learning, and relationship with others (Gunuc and Kuzu, 2015). Also, study engagement is associated with achievement in the academy and psychological well-being (Deakin Crick and Goldspink, 2014). Engagement in this study acts as a mediator that links self-regulation to behavioral intention to reuse. More explicitly, learners who can apply self-regulation strategies and skills will perceive a more positive state of passion and energy, which, in turn, leads to continuous use of e-learning systems.

The conservation of resource theory is a renowned paradigm for understanding study engagement (Gunuc and Kuzu, 2015). According to this theory (Hobfoll, 1989), engagement consumes energy, so individuals need to replenish their resources to maintain a high level (Xu et al., 2019a,b). The e-learning literature has also suggested that the e-learning environment is characterized by self-control, autonomy, and independence (Lynch and Dembo, 2004; Chiu et al., 2005); engagement in learning heavily relies on a student’s ability to control, manage, and plan learning programs (Ally, 2004). We assumed that self-regulation may help learners actively engage in their studies by adding their resources. First, self-regulation strategies could help learners acquire cognitive resources. For example, rehearsal and metacognitive strategies have been shown to be positively related to the acquisition and retention of knowledge (Yukselturk and Bulut, 2007). Second, self-regulation enables learners to manage their emotions in the pursuit of studying goals, which provides emotional resources for student engagement. Studies have shown that strategies of regulating emotions, such as situation selection, situation modification, attentional deployment, and response modulation, are linked to learners’ positive emotions (Webster and Hadwin, 2015). Therefore, it can be seen that self-regulation may facilitate student engagement.

Furthermore, study engagement can affect students’ behavioral intention to reuse e-learning systems. Engagement is characterized by dedication, absorption, and vigor. Dedication reflects individuals’ desire to invest effort in studying. Absorption refers to the state individuals enter when they are completely immersed in a task. Vigor emphasizes one’s excitement and interest in study (Schaufeli et al., 2002a; Axelson and Flick, 2010). Study engagement involves a series of activities, including behavioral engagement (e.g., concentration), emotional engagement (e.g., belonging and excitement), and cognitive engagement (e.g., mental effort; Webster and Hadwin, 2015). In terms of the conservation of resource theory, learners tend to reuse e-learning systems when they perceive them to have value. Close engagement will make learners more persistent in their e-learning. Previous research has suggested that study engagement leads to a higher return in terms of satisfaction (Webster and Hadwin, 2015); academic performance (Carini et al., 2006); student success; and retention (Wyatt, 2011). In light of the above, we proposed the following hypotheses:

Although we expected that self-regulation would positively influence study engagement, which in turn impacts behavioral intention to reuse, the conservation of resource theory (Hobfoll, 1989) indicates that environmental factors influence how learners manage and make use of self-regulation strategies and skills and obtain new resources. In e-learning settings, peer collaboration is regarded as a form of learner support where students are encouraged to exchange ideas with other students (Biasutti, 2011). It is also an important influence on mental health (Kim et al., 2020). Drawing on these insights, we proposed that peer collaboration presents an environmental resource that may strengthen the positive impact of self-regulation on study engagement.

Learners with high levels of peer collaboration tend to possess greater learning resources and are subsequently more capable of engaging with their studies. Scholars have suggested that peer collaboration facilitates both academic and social integration (Papanikolaou and Boubouka, 2010; Al-Abri et al., 2017). Learners with high levels of peer collaboration are more likely than learners with low levels of peer collaboration to perceive their self-regulation strategies and skills as an effective way to acquire more resources. Logically, learners who employ self-regulation already have the effective learning ability to obtain resources with which to invest in study engagement. If learners also receive a high level of peer support, they will become more confident in participating in e-learning activities. Peer collaboration not only helps learners transfer information and ideas to improve their self-regulation but also helps them to cooperate, improve learning climates, and enhance learning satisfaction. As a result, the relationship between self-regulation and study engagement should be stronger for learners with high levels of peer collaboration. By contrast, learners with low levels of peer collaboration should be more vulnerable to insufficient learning-related resources and less capable of investing their resources to enhance the outcomes of self-regulation, thus decreasing the effectiveness of self-regulation for integrating personal and contextual resources. Accordingly, we proposed the following hypothesis:

Given that peer collaboration moderates the relation between self-regulation and study engagement and considering that study engagement is positively associated with behavioral intention to reuse, it is logical to suggest that peer collaboration also moderates the strength of the mediating mechanism for study engagement in the relationship between self-regulation and behavioral intention to reuse, which presents a moderated mediation model (Edwards and Lambert, 2007). As previously mentioned, a stronger relation between self-regulation and behavioral intention to reuse will appear among learners with high levels of peer collaboration. Hence, the indirect effect of self-regulation on behavioral intention to reuse via study engagement may be stronger for a learner who collaborates with their peers. In particular, when a student has a high level of peer collaboration, the indirect effect of self-regulation on behavioral intention to reuse should be stronger. However, when a student is lacking such a high level of collaboration, self-regulation is less influential in promoting study engagement; consequently, the indirect effect of self-regulation on behavioral intention to reuse should be weaker. In light of the above, we proposed the following hypothesis:

We drew our sample from one large university in central China. The e-learning programs were part of the students’ regular school curriculum. The questionnaires were distributed among undergraduates within the e-learning system. They were asked to complete a self-assessment of their self-regulation, study engagement, peer collaboration, and behavioral intention to reuse the e-learning system. All participants were assured that all data would be anonymous. They signed an electronic consent form agreeing to their participation in the survey.

With the help of their professors, 400 students who engaged in e-learning activities were invited to take part. From this sample, 379 valid samples were retained a response rate of 94.75%. The average age was 20.72years (SD=2.17); 67.81% were male, and 32.19% were female. In terms of grades, 19.90% were freshmen, 33.25% were sophomores, 24.01% were juniors, 14.77% were seniors, and 8.97% were students in the fifth year. Their majors were as follows: 8.18%, humanities; 33.25%, economics and management; 15.57%, science; 54.35%, engineering; 1.85%, law; 3.96%, medicine; and 3.69%, education.

First, the confirmatory factor analysis with Mplus7.0 was used to examine the distinctiveness of the latent variables. The results showed that a theorized four-factor model distinguishing between self-regulation, study engagement, behavioral intention to reuse, and peer collaboration was a better fit to the data (χ²=482.34, df=113, p<0.001; CFI=0.94; TLI=0.93; RMSEA=0.09; SRMR=0.04) than alternative models: (a) a four-factor model in which self-regulation and study engagement were combined into one factor (χ²=1211.41, df=116, p<0.001; CFI=0.83; TLI=0.81; RMSEA=0.16; SRMR=0.07); (b) a three-factor model in which self-regulation and behavioral intention to reuse were combined into one factor (χ²=973.54, df=116, p<0.001; CFI=0.87; TLI=0.85; RMSEA=0.14; SRMR=0.07); (c) a two-factor model in which self-regulation, study engagement, and behavioral intention to reuse were combined into one factor (χ²=1542.61, df=118, p<0.001; CFI=0.78; TLI=0.75; RMSEA=0.18; SRMR=0.08); and (d) a one-factor model in which all latent variables were combined into one factor (χ²=2152.89, df=119, p<0.001; CFI=0.69; TLI=0.65; RMSEA=0.21; SRMR=0.09). The results illustrated that the participants were able to distinguish the studied variables.

Finally, we used the unmeasured latent method factor to examine the presence of common method variance (CMV; Podsakoff et al., 2003). Compared with the theorized four-factor model (χ²=482.34, df=113, p<0.001; CFI=0.94; TLI=0.93; RMSEA=0.09; SRMR=0.04), the indices of the five-factor model with all items loading on a latent common method (χ²=320.04, df=97, p<0.001; CFI=0.96; TLI=0.95; RMSEA=0.08; SRMR=0.03) were not significantly improved. The findings indicated that the data were not inflated with CMV.

The means, SDs, and correlations of the variables are provided in Table 1.

Hypothesis 1 stated that self-regulation was positively associated with behavioral intention to reuse e-learning. Hypothesis 2 further argued that study engagement mediated the relationship. The hierarchical regression analysis run by the Statistical Package for the Social Sciences (SPSS) was used to test Hypothesis 1. The mediation procedure macro developed by Preacher and Hayes (2008) was used to test Hypothesis 2. The results of hypothesis tests are summarized in Figure 2.

The direct pathway in Figure 2 shows that self-regulation was positively associated with behavioral intention to reuse (B=0.57, SE=0.05, p<0.001), controlling for socio-demographic variables. Thus, Hypothesis 1 was supported. The indirect pathway in Figure 2 further shows that self-regulation was positively related to study engagement (B=0.81, SE=0.04, p<0.001), which in turn was positively associated with behavioral intention to reuse (B=0.52, SE=0.05, p<0.001). The 95% CI for the indirect effect was [0.32, 0.52], with an average of 0.42, and did not include 0. Thus, Hypothesis 2 was supported. Additionally, Figure 2 shows that the direct effect of self-regulation on behavioral intention to reuse was still significant (B=0.15, SE=0.06, p<0.001). According to Baron and Kenny (1986), study engagement partially mediated the positive relationship between self-regulation and behavioral intention to reuse.

Hypothesis 3 stated that peer collaboration moderated the effect of self-regulation on study engagement. Hypothesis 4 further stated that peer collaboration moderated the indirect effect of self-regulation on behavioral intention to reuse via study engagement. The moderation and moderated mediation procedure macro developed by Preacher and Hayes (2008) was used to test Hypothesis 3–4. The results of hypothesis tests are summarized in Table 2.

The moderation analysis shows that the interaction of self-regulation and peer collaboration positively predicted study engagement (B=0.08, SE=0.02, p<0.01) when the controls are controlled. Table 2 shows that the simple slope test showed that simple slope was 0.44 (SE=0.06, p<0.001) at low level of peer collaboration (−1sd=−1.30), and it was 0.63 (SE=0.06, p<0.001) at high level of peer collaboration (+1sd=1.30). Figure 3 clearly illustrates the moderating role of peer collaboration in the self-regulation-study engagement association. Thus, Hypothesis 3 was supported.

Table 2 also shows that the indirect effects of self-regulation on behavioral intention to reuse via study engagement varied significantly across different levels of peer collaboration. The indirect effect of self-regulation on behavioral intention to reuse was more significant when peer collaboration was high (B=0.33, SE=0.05, p<0.001) than those when peer collaboration was low (B=0.23, SE=0.05, p<0.001). Figure 4 clearly illustrates the moderating role of peer collaboration in the indirect effect of self-regulation on behavioral intention to reuse via study engagement. Hypothesis 4 was therefore supported.

First, our research contributes to the e-learning literature by proposing and confirming the effect of self-regulation on behavioral intention to reuse an e-learning system. E-learning research has highlighted the importance of considering self-regulation strategies and skills because they influence how learners manage, plan, and reflect learning process (Jansen et al., 2019, 2020). These studies appear particularly valuable for identifying ways to improve the success of e-learning. However, in many cases, learners’ continuance usage of e-learning, such as behavioral intention to reuse, are of limited study for the outcomes of self-regulation. The current study, enriching the effects of self-regulation, indicates that self-regulation would influence behavioral intention to reuse e-learning. This is one of the limited studies that establish the relationship between self-regulation and behavioral intention to reuse.

Second, our study further contributes to the research on e-learning by identifying study engagement as a mediator linking self-regulation and behavioral intention to reuse. Given the importance of behavioral intention to reuse for e-learning success, previous studies have mostly focused on several important influencing factors (e.g., performance expectancy, effort expectancy, and social influence; Zhang et al., 2020), and what has been missing from research is study engagement. This construct represents learners’ passion and dedication to spend their time and energy on their study work. Consequently, a better understanding of the role of study engagement in influencing behavioral intention to reuse is needed. Drawing on conservation of resource theory (Hobfoll, 1989), we find that learners who can employ self-regulation strategies tend to experience emotional, cognitive, and physical engagement, and hence, it is more likely for them to reuse e-learning.

Third, by applying a moderated mediation framework, we revealed the moderating role of peer collaboration, supporting the applicability of the conservation of resource perspective in explaining the link between self-regulation and behavioral intention to reuse through study engagement. This study extends the self-regulation literature by answering the call to examine boundary conditions for the effectiveness of self-regulation. Self-regulation has been established as a strong predictor of positive outcomes of e-learning (El-Adl and Alkharusi, 2020; Jivet et al., 2020). Yet, very little research has focused on the potential boundary conditions that either promote or hinder the positive effects of self-regulation. This study suggests that high peer collaboration strengthens the positive effect of self-regulation on behavioral intention to reuse and study engagement. Learners who learn in an environment that provides peer collaboration have highly positive learning support. Thus, we can view high peer collaboration as part of a positive external environment. Our findings provide insights into how peer collaboration, as a contextual factor to facilitate self-regulation, can boost the increased intention to adopt e-learning systems.

First, for educators, more attention should be paid to self-regulation. Learners’ self-regulation has always been regarded as an essential goal perused by e-learning. Self-regulation undoubtedly increases the outcomes of e-learning; moreover, the results of this study showed that self-regulation has a positive effect on the behavioral intention to reuse of the learners. This indicates that if educators cultivate learners’ self-regulation strategies and skills, learners’ adoption of e-learning will be increased. Therefore, educators should focus on how to improve self-regulation ability among students when they want to increase their enthusiasm for reusing the e-learning system. Specifically, mental effort (van Gog et al., 2020) may be an effective way to help students to improve self-regulation in higher education.

Second, appropriate ways should be chosen to improve the study engagement of the learners. Activities can be held to increase the contextual resources of learners to make more resources, which is beneficial to the study engagement of learners. Universities are expected to increase students’ psychological influences, such as teachers and teaching practice, feeling part of a learning community, curriculum, and assessment, which greatly benefits the generation of study engagement (Kahu, 2013).

Finally, the different levels of peer collaboration of learners should be fully explored and the collaborative learning activities (Laal and Ghodsi, 2012; El Mhouti et al., 2017) and the social integration process (Blau, 1960; Severiens and Schmidt, 2009) should be taken. For example, more consideration should be given to social and psychological support (Li et al., 2018), such as the contacts between peers, cooperative work, a cooperative learning environment, and quality of teacher-student, which could increase environmental resources for learners.