“Antecedents promoting e-learner’s engagement behavior: Mediating effect of e-learner’s intention to use behavior”

Abstract

Despite the massive growth and benefits of online learning platforms, engaging and retaining learners showcases a major challenge in the present scenario. There is a dearth of literature on measuring the antecedent factors of learner engagement behavior through mediating effect in the online learning context. Therefore, the current study was designed to empirically confirm the factors influencing the learner’s engagement in online learning platforms through the Unified Theory of Acceptance and Use of Technology (UTAUT) framework. The data was collected from 336 learners who have accessed online learning platforms. Structural equation modeling was tested to evaluate the proposed relationships in the conceptual framework. The results identified performance expectancy, facilitating conditions, learner self-efficacy, and e-learner’s intention to use behavior as key antecedents of e-learner’s engagement behavior in the online learning platform. Subsequently, the e-learner’s intention to use behavior played a significant role as a mediator in influencing the e-learner’s engagement. The novelty of the study is to offer statistical significance on limitedly studied phenomena in online learning towards investigating the e-learner’s intention to use behavior as mediation in envisioning e-learner’s engagement behavior. Additionally, the study theoretically contributes to the existing body of literature. Also, the study findings underscore the need for organizations, policymakers, and academic institutions to develop online learning platforms that enhance learners’ engagement behavior.

Keywords:

• online learning
• e-Learner engagement
• learner self-efficacy
• UTAUT
• and behavioral intention to use

PUBLIC INTEREST STATEMENT

The current study utilizes the Unified Theory of Acceptance and Use of Technology Model to analyze the factors that influence learner engagement behavior and examine the mediating role of behavioral intention to use on learner engagement behavior. The data for the study analyzed using Structural Equation Modelling (SEM-Smart PLS). The findings exhibit that learner tends to adopt and engage when they perceive online learning platforms to be compatible with the requirements and adoption results in enhanced outcomes, success, and convenience. The research findings on the mediating effect of behavioral intention to use exhibits that the learner’s intention processes the learner’s evaluative judgment of ease of use, performance expectancy, self-efficacy, technical assistance, and social influence to determine engagement in the online learning platform. Based on the learner’s perception of usefulness, ease of operation, expectations of others, confidence levels, and infrastructural amenities, an individual develops a cognitive mechanism that comprises behavioral intention. The findings of the study aid the policymakers, online learning organizations, and academic institutions in developing the online learning platforms that enhance the engagement behavior of the online learners.

1. Introduction

Information and Communication Technology (ICT) and internet penetration are the sources of social and economic development and have become an indivisible unit of all sectors, including the education sector (Goh & Yang, 2021; Tsai et al., 2017). A technologically enabled environment is a suitable and supportive system for teaching, learning, and relearning new skills or knowledge (Rajabalee & Santally, 2020). Online learning platforms are considered the future of education, with a stronger presence in higher education institutions (Risdianto et al., 2021). It is used as a supplement means for independent and classroom learning. Various online learning mediums are available such as web-based learning systems, learning management systems, Massive Open Online Courses (MOOCs), etc. MOOCs are commonly used for digitally delivering educational content for credits, certificates, or degrees (Masito et al., 2021). The popularity of online courses is demonstrated by the growth in the global online learning market and is expected to reach 325 billion dollars by 2025 (Markets, 2017; McCue, 2018). The number of learners in India is anticipated to reach 13.6 million by 2023 (IBEF, 2022; Lidhoo, 2021). Thus, online learning has become a major platform for learning and training as an alternative to the traditional school system (Deka, 2021). The growth of online learning courses is mainly justified due to associative benefits such as ease of use, flexibility, low cost, and the surge in internet and smartphone users. Alongside, the COVID-19 pandemic has also seriously impacted and urged educational institutions to deliver uninterrupted service through technology adoption (Edem Adzovie & Jibril, 2022; Heo et al., 2021; Yuan et al., 2021). Online education is considered the key element for academic institutions and employers, associated with numerous benefits for stakeholders like learners, employees, parents, and facilitators across the globe (Guo et al., 2016; Hixon et al., 2012; Ma et al., 2021).

Technology-enabled learning environment assists the users in developing creative and critical thinking abilities by rearranging and reconstructing learning-related activities (Muthmainnah et al., 2022; Tsai et al., 2017). Additionally, it is widely acknowledged for equity, access, and quality in academic settings. A key advantage of online learning platforms is that they are independent of time and location. Despite the phenomenal growth of the online learning market, learners’ retention (Liu & Pu, 2020; Ray et al., 2020) and engagement continue to be of major concern to service providers and employers (Erdoğdu & Çakıroğlu, 2021; Guo et al., 2016; Rodríguez-Ardura & Meseguer-Artola, 2016; Tani et al., 2021). Furthermore, an online learning environment demands learners to be self-regulated as it lacks face-to-face or direct monitoring and attention by the instructor. Therefore, it results in discontinuance, low persistence, and engagement among online learners. Active engagement of the learner predicts the quality of the online learning platform. Creating an environment that engages the learners with high involvement levels and commitment during the learning phase is still challenging (Guo et al., 2016; Jung & Lee, 2018).

Understanding the engagement levels in online platforms is significant, as a higher level of engagement is associated with a higher completion rate of the course (Al-Bogami & Elyas, 2020; Bond et al., 2020; Deng et al., 2019; Hone & El Said, 2016; Kim et al., 2019). Earlier studies indicate the absence of a well-established model to assess learner engagement, specifically on learners’ perception in the technologically mediated platforms. Studies have posited that lower engagement is attributed to factors such as lower efficacy levels, lack of infrastructural facilities, and low peer influence. These attributes result in a lack of intention to accept online learning, which causes lower engagement levels (Guo et al., 2016; Liu & Pu, 2020; Pilli & Admiraal, 2017). The growth of online learning platforms through “interactive multimedia” has successfully gained the attention of educational researchers to understand the factors influencing learner engagement. However, learner engagement is limitedly studied in the context of online learning (Al-Bogami & Elyas, 2020; Bond et al., 2020; Coates, 2006; Deka, 2021; Dumford & Miller, 2018; Kennedy, 2020; Pilli & Admiraal, 2017). Furthermore, earlier studies affirm a lack of clarity in confirming the antecedents of learner engagement in online learning, thus creating demand for more research to examine the same (Tani et al., 2021; Wilson et al., 2019; Xerri et al., 2018).

Existing literature has focused on exploring the impact of adopting the latest technologies for the fulfillment of education and learning outcomes (Deka, 2021; Dumford & Miller, 2018; Kennedy, 2020; Pilli & Admiraal, 2017), but relatively little is known about how technology-enhanced learning affects learner engagement as most of the prior research has only explored learner engagement in traditional face-to-face learning setting (Bond et al., 2020; Carroll et al., 2019). Besides, to date, most studies have focused on adoption intention and continuance intention to use online learning platforms and engagement behavior in the context of service delivery, health care, and government. For instance, Z. Zhang et al. (2021) identified key factors that affected students’ adoption intention in the mandatory blended learning environment and reported Social Influence (SI) and facilitating conditions as the key predictors of behavioral intention. Another study by Sharma et al. (2021) analyzed and exhibited Performance Expectancy (PE), Effort Expectancy (EE), and Facilitating Conditions (FC) as the key predictors influencing to adopt mobile apps for government services. Alongside there were prior studies in information systems research have adopted various theories such as the theory of planned behavior (Sanne & Wiese, 2018), the theory of reasoned action (Sanne & Wiese, 2018), the technology acceptance model (Seetah & Hosanoo, 2021), UTAUT (Al-Nuaimi et al., 2022) and technology task fit (Huang & Wang, 2022), to evaluate the factors that influence the acceptance or usage intention. Among these, UTAUT is the synthesis of the older theories and models tested and verified by the researchers in the diverse study context. It is primarily adopted by social science researchers to examine the behavioral intention and actual behavior of individuals. Behavior intention (BI) refers to an “individual’s readiness to perform or engage in a specific behavior.” Experiencing a stronger intention to perform the behavior, the more likely the behavior is performed by the individual, which implies favorable BI to use formed by PE, EE, SI, FC, and Learner Self-Efficacy (SE), which in turn motivates the learner to engage in the online learning platform. Indeed, studies by Abbad (2021) and Z. Zhang et al. (2021) reported that perceived utility, perceived ease of use, social influence, infrastructural resources, and higher efficacy levels lead to the formation of positive intentions among e-learners. Hence, promoting favorable use intention may promote learner engagement behavior.

Despite the extensive use of the theory by various researchers, there is still very little scientific evidence in understanding the antecedents of learner engagement behavior in technology-enhanced platforms using the UTAUT framework. Moreover, existing studies have recommended (Tamilmani et al., 2021; Venkatesh et al., 2016) to enrich the theory with new endogenous, exogenous, mediating, and moderating variables. So far, studies have examined satisfaction (Gerlach et al., 2014), job offer success (Buettner, 2017), and economic development (Venkatesh et al., 2016) as the outcome variables. Limited attention has been paid to examine learner engagement behavior as an outcome variable using UTAUT.

Therefore, the current study is set out to answer the following research questions: Which are the key drivers that influence e-learner’s engagement behavior? Moreover, how does e-learner’s intention to use behavior mediate the relationship between performance expectancy, effort expectancy, social influence, facilitating conditions, learner self-efficacy, and e-learner’s engagement behavior? Based on the above research questions, the current study is designed to empirically examine the key predictors of learner engagement behavior in online learning and test the mediating effect of e-learner’s intention to use behavior between antecedents and learner engagement for the outcome of interest using the UTAUT. The study offers a deeper understanding of the antecedents of engagement in the online learning platform. The results can be used by online learning service providers and employers by proposing various means that help to predict learners’ engagement levels. Additionally, the study contributes to the existing body of knowledge by identifying and extending the UTAUT framework as a key theoretical underpinning to analyze learner engagement behavior. Furthermore, by providing empirical support for the drivers and intervening effect of the behavioral intention that influences e-learner engagement behavior that has not been studied previously as the study’s novelty.

1.1. Theoretical framework

Researchers have proposed various models to understand individuals’ technology acceptance and usage behavior (Dajani & Abu Hegleh, 2019; Duggal, 2022; Li et al., 2022). Marketers adopt various technologies and behavior-based models such as the theory of reasoned action, motivation need theory, technology acceptance model, theory of planned behavior, a framework combining the technology acceptance model, theory of planned behavior, innovation diffusion theory, and UTAUT and UTAUT 2. These models were derived from psychology, sociology, the theory of human behavior and were widely adopted to clarify the technology adoption behavior (Al-Nuaimi et al., 2022; Dajani & Abu Hegleh, 2019; Xu et al., 2022). The UTAUT model is widely accepted among market researchers to assess the determinants of intention and actual behavior as the theory is framed by combining individual, social, and organizational characteristics (Isaac et al., 2019). Learners’ retention, engagement, and adoption of new technologies are the key marketing issues (Alyoussef, 2023; Dajani & Abu Hegleh, 2019; Raffaghelli et al., 2022) that comprise intention, belief, and attitude analysis. Learners’ intention to adopt the technology is examined through technology theories, for instance, the Technology Acceptance Model (TAM) (Davis et al., 1989). However, the theory is based on the personal factors sidelining the role of social factors and facilitating conditions inducing acceptance and usage (Abdekhoda et al., 2022; Raffaghelli et al., 2022).

According to Venkatesh et al. (2003) and Venkatesh et al. (2012), scholarly work has reviewed and integrated eight theories to propose a new integrated model named UTAUT, which robustly explains the user intention and acceptance of information technology with the four core constructs, i.e., performance expectancy, effort expectancy, facilitating conditions, and social influence moderated by voluntariness, experience, gender, and age and thereby enhancing all the previous technology acceptance models. UTAUT assists the various decision-makers in evaluating the probability of success when a new technology is introduced and aids in ascertaining the significant factors that contribute to the acceptance of new technology for the desired population who are less likely to adopt and make use of it (Davis et al., 1989; Venkatesh et al., 2003). Since the inception of the UTAUT framework, researchers have widely applied the theory in diverse study contexts (Hermita et al., 2023; Osei et al., 2022; Qiao et al., 2021; M. Raza et al., 2022) and has been affirmed by Venkatesh et al. (2003) that it explains up to 74% variance in explaining the behavioral intention. However, no published work reflects on assessing the UTAUT framework on learner engagement behavior in an online learning context (Abdekhoda et al., 2022; Osei et al., 2022; Venkatesh et al., 2016). Hence, the researcher’s outcome of interest is to examine learner engagement through the UTAUT framework in elucidating intention and behavior in online learning.

1.2. Research variables, hypotheses, and model

1.2.1. Performance Expectancy (PE), e-learner’s Intention to Use Behavior (IUB), and e-learner’s Engagement Behavior (LE)

Performance expectancy denotes an individual’s perception grounded on the usefulness of technology (Qiao et al., 2021; Venkatesh et al., 2003, 2016). PE is the strongest predictor of behavioral intention (Venkatesh et al., 2003). This construct has been derived from perceived usefulness (Davis et al., 1989, 1989), extrinsic motivation (Davis et al., 1992), job fit (Thompson et al., 1991), relative advantage (Moore & Benbasat, 1991), and outcome expectations (D. R. Compeau & Higgins, 1995; D. Compeau et al., 1999). Researchers have proved the relationship between PE and IUB and denote the individual’s commitment to a specific behavior (Abbad, 2021; Alghazi et al., 2021; Arain et al., 2019; Arkorful et al., 2022; Bajaj et al., 2021; Naveed et al., 2020; Osei et al., 2022; Qiao et al., 2021,Raza et al., 2022; Yunus et al., 2021;). In consistence with the definition provided in the UTAUT model, learners who believe in the usage of the system in enhancement of experience are likely to adopt the same for the purpose of engagement and vice versa (Yang et al., 2017). In the context of our study, the construct is used to analyze the learner’s belief concerning the benefits of the online learning system. Positive perception of the attributes of usefulness has previously resulted in customer engagement (Alyoussef, 2023; McLean & Wilson, 2019; Razmerita et al., 2019). Hence, the following hypothesis is proposed to be tested.

H1:

Performance Expectancy has an influence on the e-learner’s intention to use behavior.

H1a:

Performance Expectancy has an influence on e-learner’s engagement behavior.

1.2.2. Effort Expectancy (EE), e-learner’s Intention to Use Behavior (IUB) and e-learner’s Engagement Behavior (LE)

As per the researchers, EE is defined as “the degree of ease associated with the use of the system” (Venkatesh et al., 2003, 2012). In the UTAUT model, EE is postulated to determine behavioral intention. Few researchers examined the direct relationship between EE and behavioral intention (BI) and reported that there is a significant relationship between EE and BI (Abbad, 2021; Al-Nuaimi et al., 2022; Alghazi et al., 2021; Arain et al., 2019; Arkorful et al., 2022; Bajaj et al., 2021; Chao, 2019; Ifedayo et al., 2021; Mulik et al., 2019; Naveed et al., 2020; Yunus et al., 2021). In contrast, some studies reported an insignificant relationship between EE and BI (Fianu et al., 2018; Khechine et al., 2020; Tseng et al., 2019; Yang et al., 2017). Despite this fact, there was evidence examining the mediating effect of PE between EE and BI (Briz-Ponce et al., 2017; Rahi & Ghani, 2019). The significance of the relationship between EE and BI is still debatable (Qiao et al., 2021). Learners are likely to adopt and engage in the system if they believe a system is easy to use or operate (McLean & Wilson, 2019; Osei et al., 2022). Furthermore, the engagement levels of the learners depend on the user-friendliness of the online learning platforms (Sharma et al., 2018). Hence, the following hypothesis is proposed:

H2:

Effort Expectancy has an influence on the e-learner’s intention to use behavior.

H2a:

Effort Expectancy has an influence on e-learner’s engagement behavior.

1.2.3. Social Influence (SI), e-Learner’s Intention to Use Behavior (IUB), and e-learner’s Engagement Behavior (LE)

Social Influence refers to the external social environment associated with an individual’s behavioral intention, such as reflection of friends, instructors, subjective norms, and peers (Qiao et al., 2021; Venkatesh et al., 2016). Subjective norm is the origin of this construct (I. Ajzen, 1991; M. Ajzen et al., 1975; Davis et al., 1989; Mathieson, 1991; Taylor & Todd, 1995), Social factors (Thompson et al., 1991), Image (Moore & Benbasat, 1991). Learners’ intention to adopt and use an online learning platform normally depends on the suggestion of friends, instructors, or superiors. Literature has found a direct association between social influence and behavioral intention on technology adoption (Qiao et al., 2021). SI creates pressure for the learners to adopt the behavior of their peers. The stronger the effect of the SI, the greater is the occurrence of a behavior. Earlier studies have proved a strong association between SI and BI (Al-Nuaimi et al., 2022; Arain et al., 2019; Chopdar & Sivakumar, 2018; Hermita et al., 2023; Ifedayo et al., 2021; Naveed et al., 2020; Yeboah & Nyagorme, 2022; Yunus et al., 2021; Zhang et al., 2021), which indirectly affects the actual behavior. On the contrary, authors have expressed that SI and normative practices enforce individuals to enhance engagement (Abbad, 2021; Fu et al., 2020; McLean & Wilson, 2019; Pilli & Admiraal, 2017; Razmerita et al., 2019). Therefore, it is hypothesized that.

H3:

Social Influence has an influence on the e-learner’s intention to use behavior.

H3a:

Social Influence has an influence on the e-learner’s engagement behavior.

1.2.4. Facilitating Condition (FC), e-learner’s Intention to Use Behavior (IUB), and e-learner’s Engagement behavior (LE)

Facilitating Condition refers to the availability of technical support and resources to adopt the technology (Khechine et al., 2020; Qiao et al., 2021; Venkatesh et al., 2003). The construct is adopted from perceived behavioral control (I. Ajzen, 1991; Taylor & Todd, 1995), facilitating condition (Thompson et al., 1991) and compatibility (Moore & Benbasat, 1991). External resources provided to the learners in the form of infrastructure and support to use online learning systems result in a positive attitude towards usage behavior (Al-Nuaimi et al., 2022). Studies reported a significant influence of FC on intention and usage behavior (Rodríguez-Ardura & Meseguer-Artola, 2016; Fianu et al., 2018; Arain et al., 2019; Mulik et al., 2019; Naveed et al., 2020; Camilleri, 2019; Tseng et al., 2019; Khechine et al., 2020, Zhang et al., 2020; Abbad, 2021; Ifedayo et al., 2021; Yunus et al., 2021; Abdekhoda et al., 2022; Al-Adwan et al., 2022). Likewise, efficient adoption and engagement of the learner in the online learning platform majorly rely on the infrastructure facilitated for online learning (McLean & Wilson, 2019). Hence, the hypothesis postulated is as follows.

H4:

Facilitating Condition has an influence on e-learner’s intention to use behavior.

H4a:

Facilitating Condition has an influence on e-learner’s engagement behavior.

1.2.5. Self-Efficacy (SE), e-learner’s Intention to Use Behavior (IUB), and e-learner’s Engagement behavior (LE)

Self-Efficacy is defined as the “measurement of the degree or strength of an individual’s belief in the ability to complete tasks and achieve goals”. Successful adoption of online learning systems is strongly associated with the learner’s motivation and confidence in utilizing the technology (Chiu & Wang, 2008; D. Compeau et al., 1999; Qiao et al., 2021; Wu & Tsai, 2006; K. Zhang & Yu, 2022). Individuals with optimistic self-efficacy are likely to have positive views of technology and frequently intend to use technology (Schaupp, Carter & Mcbride, 2010). Chiu and Wang (2008) found computer self-efficacy to be the significant predictor of an individual’s intention to continue to use the web-based learning system. Various researchers have investigated the role of SE in technology adoption in diverse contexts. Few studies have reported SE as significant in determining behavioral intention (Schaupp et al., 2010; Fianu et al., 2018; Han & Yi, 2018; Hermita et al., 2023; McKenna et al., 2017; Raffaghelli et al., 2022; Xu et al., 2022). Learners with higher efficacy levels are intended to have greater engagement levels. In this regard, SE is regarded as the crucial factor affecting the intention and engagement behavior of the learner while using the online learning platform. Furthermore, Uden et al. (2014) reported that self-efficacy levels significantly influenced learners’ emotional, cognitive, and behavioral engagement level. Studies report that a positive attitude and higher self-efficacy levels to adopt online learning platforms yield better learning outcomes, satisfaction, and enhanced engagement levels (Razmerita et al., 2019; Sun & Rueda, 2012). Therefore, the current study tests the proposed hypothesis:

H5:

Self-Efficacy has an influence on e-learner’s intention to use behavior.

H5a:

Self-Efficacy has an influence on the e-learner’s engagement behavior.

1.2.6. E-learner’s Intention to Use Behavior (IUB) and e-learner’s Engagement behavior (LE)

IUB is widely validated as the key antecedent of the behavior and is conceptualized in the UTAUT model as the key predictor of the actual behavior (Venkatesh et al., 2003, 2012). IUB is the effort an individual is prepared to exert to perform the behavior. Different behavioral theories are widely applied to evaluate and predict the engagement levels in the diverse study context. Sanne and Wiese (2018) adopted the theory of planned behavior to ascertain the user engagement level in Facebook advertising. Engagement is manifested as the behavioral outcome of an individual when exposed to a stimulus (Sanne & Wiese, 2018). The level of social engagement encouraged in online courses enhances retention behavior (Aldowah et al., 2020). The user’s actual engagement was significantly influenced by the user’s behavioral intention. The engagement levels of an individual are associated with the retention levels. The higher the engagement levels, the lower the dropout. Intention to use the online learning platform positively influenced learner engagement with the online learning platform. Fianu et al. (2018), Tseng et al. (2019), and Khechine et al. (2020) proved IUB as the significant predictor of actual behavior. Hence, the hypothesis proposed is,

H6:

E-learner’s intention to use behavior has a significant influence on the e-learner’s engagement behavior.

1.2.7. E-learner’s Intention to Use Behavior (IUB) as mediating variable

IUB is the crucial antecedent of the actual behavior, as the intention denotes the cognitive readiness of an individual to take a specific course of action (Yi et al., 2016). Behavioral intention significantly influences the engagement of the learners. Previous studies show that learners with a lower intention to adopt online learning platforms tend to show lower engagement levels. In contrast, high engagement levels are manifested by the learners with a stronger intention to adopt the learning platform (Qiao et al., 2021). Ifedayo et al. (2021) expressed that behavioral intention significantly partially mediated the relationship between performance expectancy, effort expectancy, social influence, facilitating conditions, and actual behavior. Additionally, Mafabi et al. (2017) examined the mediating role of intention in knowledge-sharing behavior. Behavioral intention fully mediated the relationship between subjective norm, perceived behavioral control, attitude, and knowledge-sharing behavior. Findings of Osei et al. (2022) reported intention as the core mediator between intention and behavior. Limited studies have examined the mediating effect of Intention to use Behavior adopting UTAUT framework in online learning context. Therefore, the proposed study intends to examine e-learner’s intention to use behavior as a key mediator between the study antecedents and learner engagement behavior. Hence, the hypothesis proposed is as follows (Figure 1):

Figure 1. Research model.

H7:

E-learner’s intention to use behavior significantly mediates the relationship between the independent variables/antecedents and e-learner’s engagement behavior, the outcome of interest.

2. Materials and methods

Research on the technology acceptance paradigm has widely adopted a quantitative approach (Ameen et al., 2019; Davis et al., 1989). For the current study, the researcher adopted quantitative research design, and the data was collected at a single point of time; hence, the study was cross-sectional in nature.

2.1. Questionnaire and measurement items

To examine the antecedents of e-learner’s engagement behavior, an online self-administered questionnaire was used to collect data. The questionnaire comprised two sections, namely, demographic factors and constructs of the proposed model and their measurement items. All the measurement scales for the study were adapted from the existing literature. All the constructs were measured using multiple items of a seven-point Likert scale ranging between “strongly disagree” and “strongly agree.” The measurement items for the construct’s Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Condition, and E-learner’s Intention to Use Behavior were directly adopted from Venkatesh et al. (2003) & (Venkatesh et al., 2012); and were measured retrospectively. Measurement items for Learner Self-Efficacy were drawn from D. R. Compeau and Higgins (1995) study. E-learner’s Engagement Behavior was measured through the scale proposed by Deng et al. (2019). A detail of the measurement items of the constructs is exhibited in Table 1.

Table 1. Measurement items

Constructs	Measurement Variable	Sources
Performance Expectancy (PE)	PE1: I find E-learning useful for meeting my learning requirements. PE2: Using an E-learning platform enables oneself to accomplish learning quickly. PE3: Using E-learning increases my chances of attaining things that are important to me.	Venkatesh et al. (2003) & (Venkatesh et al., 2012);
Effort Expectancy (EE)	EE1: It is easy for me to become skilled at using E-learning platform. EE2: My interaction with the E-learning platform is clear and understandable. EE3: Learning to work with an E-learning platform is easy for me.
Social Influence (SI)	SI1: People who influence my behavior think that I should use E-learning platform. (People here refer to: lecturers, instructors, colleagues, friends, family members). SI2: People who are important to me think that I should use an E-learning platform. SI3: People whose opinion I value prefer that I use E-learning platform.
Facilitating Conditions (FC)	FC1: I have the necessary resources to use an E-learning platform. FC2: I have the knowledge that is necessary to use an E-learning platform. FC3: Necessary guidance was provided to me when I faced difficulty to use E-learning platform. FC4: The technology which is necessary for adoption of E-learning platform in my university is updated and modern.
Learner Self-Efficacy (SE)	SE1: I could complete my learning activities using an e-learning platform if I have never used a platform like it before. SE2: I could complete my learning activities using an e-learning platform if I have only software manuals for reference. SE3: I am confident of using the e-learning platform even if there is no one around to show me how to use it.
E-learner’s Intention to Use Behavior (IUB)	IUB1: I intend to use E-learning platform in the future. IUB2: I intend to engage in E-learning routinely. IUB3: I intend to use E-learning for upgrading my knowledge and skills.
E-learner’s Engagement Behavior (LE)	LE1: I set aside a regular time each week to work on the e-learning platform. LE2: I took notes while studying on an e-learning platform. LE3: I revisited my notes when preparing for e-learning related assessment tasks. LE4: I often searched for further information when I encountered something in the e-learning that puzzled me. LE5: When I had trouble understanding a concept or an example in e-learning, I went over it again until I understood it. LE6: If I watched a video lecture that I did not understand at first, I would watch it again to make sure I understood the content. LE7: I was inspired to expand my knowledge through an e-learning platform. LE8: I found e-learning interesting. LE9: I enjoyed learning through the e-learning platform. LE10: I often responded to other learners’ questions. LE11: I contributed regularly to course discussions. LE12: I shared learning materials (e.g., notes, multimedia, links) with other classmates in the e-learning platform.	Deng et al. (2019)

Before administering the final questionnaire, the validity of the questionnaire was ascertained to check the suitability in the current study setting. Sekaran and Bougie (2016) described validity for any questionnaire as “the extent to which the measure behaves in a way consistent with theoretical hypotheses and represents how well scores on the instrument indicate the theoretical construct.” The validity of the questionnaire was examined through content validity (CV). The suitability of items in the study was obtained by approaching four experts. The expert panel comprised professors, industrialists, and statisticians with prior research experience. Experts not only supported in conducting content validity but also in refining the items for final use. After incorporating the modifications, the questionnaire was pilot tested with a sample of 40 (Hertzog, 2008) to ensure clarity, accuracy, and readability before data collection from the targeted respondents. Reliability was ensured using Cronbach Alpha value for the items using Jamovi software. The Cronbach’s alpha coefficients for the constructs exceeded the threshold value of 0.70. Based on the results of the pilot study. The questionnaire was administered for the final data collection.

2.2. Sampling and data collection

Students are the current study sample, who have access to e-learning platforms at university level. The Indian universities have created a separate provision for mandate e-learning through regulatory agencies as part of Digital India programme. So, learners from the universities have access and exposure to online learning platforms for boundaryless knowledge enrichment. The study adopted online survey method as part of convenience sampling technique. There were studies suggesting to opt convenient sampling for online surveys as it gives ease of operation, relatively inexpensive and decreases the interference of investigators while filling out the survey (Al-Adwan et al., 2022).

The sample was chosen for the study comprised both undergraduate and postgraduate students. To calculate the feasible sample size for the study, researcher utilized G*power 3.1.9.4 software (Faul et al., 2009). The minimum recommended sample size for the study was 305 (effect size f2 = 0.07, power = 0.95, α err prob = 0.05, number of predictors = 6). Additionally, the sample size was validated as per the sample determination method proposed by Hair (2017) in which each item in the survey instrument can consider a minimum of 5 to 10 respondents as the total sample size. Hence, based on the items in the finalized questionnaire, the responses are collected in the proportion of a 1:10 item-to-respondent ratio to support the study sample (Hair, 2017). Therefore, the threshold level ascertained using both methods was maintained.

The final version of the questionnaire was administered to 650 respondents through email between June 15 to 31 July 2021. To reduce the potential biases, the construct items were carefully designed with a cover letter emphasizing that partaking in the survey was voluntary and the responses would be utilized only for the research purpose and treated in strict confidence. In other words, the survey was designed to ensure that respondents remained anonymous and could withdraw when they wanted. Informed consent was given by respondents that showed they agreed to a set of statements included in the questionnaire’s cover letter. Follow-up messages were sent to reduce non-response bias. Participation was voluntary, and no incentive was offered for participation.

2.3. Data cleaning

Among 650 questionnaires circulated, 378 responses were returned. Since the survey was administered online, all the questions were mandatory, and the chances of incomplete questionnaires were nil. Researchers manually verified the questionnaire and eliminated 42 responses due to unengaged and random responses to each item within the scale. The researchers retained 336 respondents’ data for further analysis, exhibiting a response rate of 51.70%.

2.4. Data analysis

The data collected were analyzed using structural equation modeling (PLS-SEM) with SmartPLSV3.0 software. Partial Least Squares (PLS) regression is a widely accepted Structural Equation Modeling (SEM) method to validate structured data (Chao, 2019; Leguina, 2015). PLS regression is commonly used for data analysis during the initial stages of theory development and for validating the existing theory. We analyzed the quantitative data in the two stages consistent with existing literature: First, testing the measurement model specifying the measurement of latent variables in terms of the observed variables and the structural model that specifies the causal relationship between independent and dependent variables be tested. The validity and reliability of the measurement model were estimated using indices such as Cronbach Alpha, AVE, Fornell Larcker criteria, and composite reliability. After finalizing the measurement model, PLS regression was used to execute bootstrapping and validate the test for the conceptual model and examine the relationships among the hypothesized constructs.

2.5. Ethical compliance

The study adhered to research ethics governed by the Institutional Ethics Committee requirements. Respondents were made aware of the objective, benefits, and outcome of the study. Participation in the study was voluntary, and respondents were notified of their right to withdraw from the study at any time. Respondents’ anonymity was maintained throughout the research. Responses obtained from participants were kept confidential and used exclusively for academic purpose, and were communicated to the study’s target population.

3. Results

Table 2 lists the demographic information for the sample group. There were 141 males (41.96%) and 195 females (58.04%). More than half (60.42%) of the learners were in the age range between 20 and below 23, followed by below 20 (22.02%) and 23 and above (17.56%). Most of the respondents were in the second year (54.76%) of their studies. Participating learners were from the following course specialization: M.Com (30.95%); BBM (30.36%); B.Com (23.51%), and MBA (15.18%). Most of the learners had extensive e-learning experience ranging from less than a year to 6 years (92.26%).

Table 2. Demographic profile

Demographics		N	Percentage
Age	Below 20	74	22.02
	Between 20 and Below 23	203	60.42
	23 and above.	59	17.56
Gender	Male	141	41.96
	Female	195	58.04
Year of Study	First-year	96	28.57
	Second year	184	54.76
	Third year	56	16.67
Course Specialisation	B. Com	79	23.51
	BBM	102	30.36
	M.Com	104	30.95
	MBA	51	15.18
Prior Experience	Less than one year	66	19.64
	1–2 years	106	31.55
	3–4 years	95	28.27
	5–6 years	43	12.80
	7–8 years	3	0.89
	8 years and above	1	0.30
	Not at all	22	6.55

Source: Author’s computation.

The measurement model was evaluated by assessing the Convergent Validity (CV), Discriminant Validity (DV), and internal reliability. Average Variance Extracted (AVE) was used to assess the CV. Cronbach Alpha and composite reliability values for all constructs were ascertained to examine the internal reliability. Commonly used evaluation indicators (Chao, 2019; Hair et al., 2019) were chosen, such as composite reliability, AVE, and Cronbach Alpha results are shown in Table 3.

Table 3. Measurement model

Construct	Items	Mean	Factor Loading	Cronbach’s Alpha	AVE	Composite Reliability
Performance Expectancy	PE1	4.664	0.924	0.926	0.872	0.953
	PE2	4.708	0.941
	PE3	4.833	0.935
Effort Expectancy	EE1	4.801	0.889	0.871	0.794	0.920
	EE2	4.967	0.917
	EE3	5.080	0.866
Social Influence	SI1	4.821	0.806	0.851	0.769	0.909
	SI2	4.580	0.917
	SI3	4.560	0.903
Facilitating Condition	FC1	4.958	0.847	0.867	0.714	0.909
	FC2	5.030	0.872
	FC3	4.753	0.836
	FC4	4.839	0.825
Self-Efficacy	SE1	4.342	0.900	0.891	0.820	0.932
	SE2	4.494	0.926
	SE3	4.714	0.891
E-learner’s Intention to Use Behavior	IUB1	4.943	0.913	0.898	0.831	0.936
	IUB2	4.893	0.924
	IUB3	5.042	0.898
E-learner’s Engagement Behavior	LE 1	4.229	0.736	0.927	0.628	0.938
	LE 2	4.497	0.795
	LE 3	4.577	0.825
	LE 4	4.768	0.835
	LE 5	4.878	0.799
	LE 6	4.902	0.780
	LE 7	4.783	0.736
	LE 8	4.765	0.821
	LE 9	4.631	0.799
	LE 10*	4.262	0.589
	LE 11*	4.208	0.606
	LE 12*	4.229	0.548

Source: Author’s computation. *Notes: The factor loading for LE10, LE11 and LE12 ranged below the threshold i.e., 0.70; hence, three items of the LE were removed and excluded for further analysis.

Most of the factor loadings for the constructs above 0.70 (Hair et al., 2019) were considered. The factor loading for three items in the e-learner engagement construct ranged below the threshold; hence, three items of the construct were removed and excluded for further analysis. Cronbach Alpha and composite reliability are used to evaluate construct reliability which indicates how well the items measure the construct. Cronbach Alpha values for all constructs ranged between 0.927 and 0.851 and composite reliability between 0.919 and 0.953, exceeding the threshold limit of 0.7 (Hair et al., 2019), thereby exhibiting high reliability. AVE values ranged from 0.628 to 0.872 and were greater than 0.5 for every construct (Hair et al., 2019), presenting convergent validity.

To assess the discriminant validity, the square root of the AVE of each latent construct was equated with its inter-construct correlation. Acceptable DV is achieved when the square root of the AVE of a construct is greater than its correlation with other constructs (Hair et al., 2019). In addition, the diagonal values should be higher than the off-diagonal values in corresponding rows and columns (Chao, 2019). As presented in Table 4, DV was supported as the square root of AVE for the construct was more significant than the correlation with other constructs.

Table 4. Discriminant validity: Fornell and Larcker

Construct	1	2	3	4	5	6	7
E-learner’s Intention to Use Behavior (1)	0.910
Effort Expectancy (2)	0.60	0.891
Facilitating Condition (3)	0.59	0.624	0.84
E-learner’s Engagement Behavior (4)	0.55	0.532	0.55	0.792
Performance Expectancy (5)	0.58	0.549	0.59	0.585	0.934
Self-efficacy (6)	0.56	0.525	0.47	0.524	0.582	0.906
Social Influence (7)	0.44	0.572	0.59	0.448	0.459	0.478	0.877

Source: Author’s computation.

Fornell–Larcker criteria insufficiently detect discriminant validity problems (Al-Fraihat et al., 2020); hence, Heterotrait–Monotrait Ratio criteria proposed by Henseler et al. (2015) was used to assess DV. HTMT values above 0.85 question the DV. As exhibited in Table 5, all the values are below the threshold level. Therefore, it is evident that there is significant discriminant validity in the construct studied.

Table 5. Discriminant validity—HTMT

Construct	1	2	3	4	5	6	7
E-learner’s Intention to Use Behavior (1)
Effort Expectancy (2)	0.681
Facilitating Condition (3)	0.668	0.708
E-Learner’s Engagement Behavior (4)	0.603	0.584	0.617
Performance Expectancy (5)	0.635	0.595	0.665	0.628
Self-Efficacy (6)	0.620	0.588	0.527	0.570	0.640
Social Influence (7)	0.498	0.671	0.681	0.494	0.487	0.538

Source: Author’s computation.

Table 6 summarizes the structural model, and results of hypothesis, hypothesis H1, H1a, H2, H4, H4a, H5, H5a, and H6 were supported by the empirical data. While H2a, H3, and H3a were rejected. The results showed performance expectancy, effort expectancy, facilitating condition, self-efficacy has a significant influence on e-learner’s intention to use behavior (β = 0.184, P < 0.05), (β = 0.245, P < 0.001), (β = 0.263, P < 0.001), (β = 0.231, P < 0.05), (β = 0.162, P < 0.05), (β = 0.229, P < 0.05), (β = 0.152, P < 0.05), (β = 0.146, P < 0.10), respectively. However, social influence had an insignificant effect on e-learner’s intention to use behavior (β=-0.035, P = 0.579) and e-learner’s engagement behavior (β=-0.042, P = 0.507). The results showed that performance expectancy, facilitating conditions, self-efficacy, and e-learner’s intention to use behavior have a significant influence on e-learner’s engagement behavior (β = 0.245, P < 0.001), (β = 0.162, P < 0.001), (β = 0.152, P < 0.05) (β = 0.146, P < 0.10). Finally, effort expectancy and social influence had an insignificant effect on learner engagement; hence, H2a and H3 are rejected. The study results exhibit that PE, FC, SE, and IUB have a significant and positive impact on the LE, validating H1, H1a, H2, H4, H4a, H5, H5a, and H6 and providing strong empirical evidence. Therefore, the e-learner intention to use and engagement behavior is enabled by the performance expectations, enabling conditions, learner self-efficacy, and IUB. The R² value for e-learner’s intention to use behavior and e-learner’s engagement behavior ranged between 0.468 and 0.519. In terms of the variance explained, the research model accounted for 46.8% of the variance in e-learner’s intention to use behavior and 51.9% variance in e-learner’s engagement behavior. Therefore, these constructs reflect moderate predictability. The quality of the structural model is assessed using the Q² value that predicts the relevance of the structural model (Hair et al., 2019). Q² analysis is applied to the dependent variables that denote reflective measurement. Values greater than zero demonstrate that the model has adequate predictive relevance for the dependent variable. The model has adequate predictive relevance as the Q2 values for e-learner’s intention to use behavior and e-learner’s engagement behavior were 0.420 and 0.281, respectively. The research provides robust empirical evidence by identifying the key antecedents that promote the e-learner engagement behavior. Besides, though the research reproduces UTAUT connection between PE, EE, SI, FC, and IUB. It proposes new evidence through inclusion of a new construct, self-efficacy, stresses on the importance of predicting e-learner intention to use behavior and e-learner engagement behavior.

Table 6. Summary of results – Hypothesis testing

Path	Path Coefficients	t values	p values	R^2	Q^2
Performance Expectancy -> E-learner’s Intention to Use Behavior.	0.184	2.746	0.006*	0.519	0.420
Performance Expectancy -> E-learner’s Engagement Behavior	0.245	3.562	0.000***
Effort Expectancy -> E-learner’s Intention to Use Behavior.	0.263	3.594	0.000***
Effort Expectancy -> E-learner’s Engagement Behavior	0.103	1.454	0.146
Social Influence -> E-learner’s Intention to Use Behavior.	−0.035	0.556	0.579
Social Influence -> E-learner’s Engagement Behavior	0.042	0.663	0.507
Facilitating Conditions -> E-learner’s Intention to Use Behavior.	0.231	3.349	0.001*	0.468	0.281
Facilitating Conditions -> E-learner’s Engagement Behavior	0.162	2.518	0.012*
Self-efficacy -> E-learner’s Intention to Use Behavior.	0.229	2.966	0.003*
Self-efficacy -> E-learner’s Engagement Behavior.	0.152	2.325	0.020*
E-learner’s Intention to Use Behavior -> E-learner’s Engagement Behavior	0.146	1.838	0.063**

Source: Author’s computation. Notes ***p < 0.001, *p < 0.05, **p < 0.10.

The study examined the mediating effect of the intention to use behavior between PE, EE, SI, FC, SE, and learner engagement. The results in Table 7 represent the summary of the indirect effect of PE, EE, SI, FC, SE, and LE on online learning when mediated through e-learner’s intention to use behavior at p < 0.05. Mediation hypotheses were tested based on Hair et al. (2017), the assertion that if the values for the standardized total effect and standardized direct effect of a predictor variable differ, mediation is present. The findings of the study indicate the full mediation effect of intention to use behavior on the relationships between PE, EE, FC, SE, and partial mediation between SI and LE. Specifically, the indirect effects of intention to use behavior reduced the total effects to a non-significant level, thereby supporting H7. This implies that intention to use behavior plays a significant role in the processing of PE, EE, FC, and SE in influencing the engagement behavior of the learner. The study contributes to the body of literature by studying e-learner intention to use behavior (IUB) as the mediator in the UTAUT framework. Assessing IUB as a mediator enabled a clear understanding of the association between the key predictors and e-learner engagement behavior and design the effective strategies that promote the engagement levels of the learners.

Table 7. Mediating effect of E-learner’s intention to use behavior

	Performance Expectancy	Effort Expectancy	Social Influence	Facilitating Condition	Self-Efficacy	E-learner’s Intention to Use Behavior
Standardised Total Effects: E-learner’s Intention to Use Behavior. E-learner’s Engagement Behavior.	.184 0.272	.263 0.141	−0.035 0.036	0.231 0.195	0.229 0.185	0.000 0.146
Standardised Direct Effects: E-learner’s Intention to Use Behavior. E-Learner’s Engagement Behavior.	.184 0.245	.263 0.103	−0.035 0.042	0.231 0.162	0.229 0.152	0.000 0.146
Standardised Indirect Effects: E-learner’s Intention to Use Behavior. E-learner’s Engagement Behavior.	.000 0.027	.000 0.038	0.000 −0.006	0.000 0.033	0.000 0.033	0.000 0.000

Source: Author’s computation. Notes p<0.05.

4. Discussion

The present study aimed to examine the key predictors of e-learner’s engagement in online learning platforms with the e-learner’s intention to use behavior as the mediating factor of behavior between the study constructs. The UTAUT framework was adopted as the guiding framework for structural equation modeling. Based on the empirical evidence of this study, it can be inferred that to predict positive intention, the learners should be exposed to PE, EE, FC, and SE but when the e-learner’s intention to use behavior is treated as a mediator apart from SI, the remaining constructs have significantly contributed to effective e-learner engagement behavior. The results of the study support performance expectancy have a significant and direct effect on e-learner intention to use behavior and e-learner engagement behavior; the finding concurred with McLean and Wilson (2019); S. A. Raza et al. (2020); Bajaj et al. (2021); Qiao et al. (2021); Arkorful et al. (2022); Yeboah and Nyagorme (2022). Learners tend to adopt and engage when they perceive online learning platforms to be compatible with the requirements, resulting in enhanced outcomes, success, and convenience (Al-Nuaimi et al., 2022; Muhammad et al., 2021).

Consequently, the findings indicate that EE has a significant effect on e-learner intention to use behavior but does not affect e-learner engagement behavior. Instead, its effect on e-learner engagement behavior was fully mediated through e-learner intention to use behavior. This finding is in line with Mafabi et al. (2017); Ifedayo et al. (2021) and Al-Nuaimi et al. (2022). Ease of use and user-friendliness drive the intention of the learner to adopt the online learning platforms. The result shows an insignificant relationship between social influence, e-learner intention to use, and e-learner engagement behavior. This implies that even if a learner adheres to social influence, the behavioral factors do not transform into e-learner intention to use and e-learner engagement behavior. The role of social influence in explaining behavior is often complex and differs across contexts (Venkatesh et al., 2003). This study’s findings are consistent with Mafabi et al. (2017) and Rokhman et al. (2022), who posited that subjective norm insignificantly influenced behavioral intention to use e-learning. However, research findings contradict the studies by McLean and Wilson (2019); Tseng et al. (2019); Khechine et al. (2020); Ifedayo et al. (2021); Muhammad et al. (2021) and Abdekhoda et al. (2022). The finding associated with insignificant relationship implies that the support of a social circle is not required to be driven to intend to use and engage in online learning platforms for learners who have a greater task-technology fit and prioritize their own beliefs, inclinations, and values (Chatterjee & Bhattacharjee, 2020; Fianu et al., 2018; Jain et al., 2022).

Similar to the prior research findings of Fianu et al. (2018); McLean and Wilson (2019); Mulik et al. (2019); Tseng et al. (2019); Khechine et al. (2020); Ifedayo et al. (2021); Muhammad et al. (2021); Yeboah and Nyagorme (2022) facilitating condition was found to have a significant influence on e-learner’s intention to use and e-learner’s engagement behavior. Limited access to the resource prevents the learners from engaging in the system (Hermita et al., 2023; Hoi, 2020). Therefore, resources in the form of good network connectivity, higher browsing speed, and elimination of bandwidth problems during the learning process fosters the engagement levels of the learners. Our findings also indicate that learner self-efficacy has a significant and direct effect on the e-learner intention to use and e-learner engagement behavior, demonstrating that individuals with higher self-efficacy tend to show positive intention to use and higher engagement levels (Heo et al., 2021). The study results are in confirmation with findings of other studies (Ameen et al., 2019; Fianu et al., 2018; Hoi, 2020; Li et al., 2022; K. Zhang & Yu, 2022) emphasizing on the self-regulated behavior of the learners for favorable intention to use behavior and engagement behavior. Therefore, it is crucial to foster learners’ confidence in using the online platform by providing technical support and training. This research provides empirical evidence that the intention to use behavior directly affects learner engagement. The intensity of the intention drives the engagement of the learner. The findings are in line with Sanne and Wiese (2018), who found that the user’s intention to use behavior significantly influenced user engagement in Facebook advertising. Although the study findings are proven in the context of social media advertising, these findings provide valuable insights to the tech-based organizations and service providers of online learning platforms to consider it as a key antecedent for enhancing the engagement levels.

To explicate the mediating effects in the study, the research findings empirically provide evidence that e-learner’s intention to use behavior fully mediates the relationships between PE, EE, FC, SE, and partially mediates the relationship between SI and e-learner’s engagement behavior. This finding demonstrates that the learner’s perception of usefulness, ease of operation, expectations of others, confidence levels, and infrastructural amenities, an individual develops a cognitive mechanism that comprises intention to use and engagement behavior. The finding concurred with Mafabi et al. (2017); Hoi (2020) and Ifedayo et al. (2021).

Theoretically, the novelty of the current research lies in the examination of key antecedents of learner engagement behavior in online learning platforms. Prior studies on UTAUT have majorly focused on the adoption of online technologies. However, limited studies have explicitly analyzed e-learner engagement behavior as an outcome variable. By studying e-learner engagement as an outcome variable, researchers gain a deeper understanding of how technology aspects, IUB, EE, SE, and SI, impact the engagement levels of the learners. Furthermore, the inclusion of a new construct, self-efficacy, emphasized the importance of predicting e-learner intention to use behavior and e-learner engagement behavior. Additionally, the novelty of the study rests in studying e-learner intention to use behavior (IUB) as the mediator in the UTAUT framework. Examining IUB as a mediator enabled a clear understanding of the association between the key antecedents and e-learner engagement behavior and design the effective interventions that promote the engagement levels of the learners. Overall, the study contributes to the existing body of knowledge by providing empirical evidence on the relationship between the study constructs in the online learning context and highlighting the predominant factors that promote positive learner engagement behavior.

4.1. Implications for further research and practice

Since most of the institutions use online learning platforms and are investing a significant amount for effective use and facilitating the learning process (Al-Fraihat et al., 2020), the present study responds to the scholarly calls for future research on the understanding of antecedents and behavioral outcomes of the intention in the online learning platforms. Researchers advanced the UTAUT theoretical model by examining e-learner engagement behavior as the outcome variable. The results of the present study add support to the scholarly literature on learner engagement in online courses by validating a model encompassing technological factors (facilitating conditions, PEOU), individual factors (learner self-efficacy), and communal factors (social influence) that influence the e-learner intention to use behavior to use technology and learner engagement behavior.

The study’s results provide empirical evidence that positive evaluation of the behavior, resources, community influences, and competence exert a positive effect on the intention to use online learning platforms that lead to engagement behavior. This study proffers an integrative understanding of the antecedents that influence learners to engage in the digital learning platforms actively. Furthermore, researchers extend the UTAUT framework by adding learner self-efficacy as an independent variable. The inclusion of learner self-efficacy in the model supports the relationship between e-learner intention to use behavior and learner engagement. The extended model reported a 46.8% variance in the research model. The study suggests that the variables considered for the study significantly improved the prediction ability of the proposed model and that are essential for enhancing learner engagement in online learning platforms.

Similarly, research extends the extant literature by investigating e-learner intention to use behavior as the mediator in the online learning context. The research findings empirically confirm that e-learner intention to use behavior fully mediated the relationships between PE, EE, FC, and SE and partially mediated the relationship between SI and e-learner engagement behavior. This significant contribution paves the way for further research to analyze the mediating impact of intention by introducing additional variables.

For the effective use of online learning platforms, the systems should be positioned in a way that drives the user’s intention and enhances the engagement behavior of the learners. The study results provide novel insights on the critical issues and recommendations that should be considered for improving the perception of intention and learner engagement while using online learning platforms. Due to the concern that most institutions rely on online courses, the study results emphasize the need to systematically evaluate learners’ intention for persistent improvements and address the problems and underperformances. Therefore, more effort should be directed to effectively utilize tools to exploit online learning to the fullest capacity. The outcome of the investigation points out two major sources for enhancing the favorable intention, which leads to learner engagement. PE and EE continue to be core indicators in the digital learning environment. A flexible and compatible online learning platform plays a vital role in aiding learners to achieve their academic goals. The results of the study reveal that social influence no longer determines the e-learner’s intention to use and e-learner engagement behavior. Rather, resources, perceived benefits, technology accessibility, and perceived ease of use are of greatest importance for the learners. These findings direct the instructors and course designers to increase awareness among the learners about the usefulness and benefits of online learning platforms through arranging workshops and training.

The enrollment levels of learners in online learning platforms have consistently improved over the years globally. The application of technology to education in developing countries is still at a very nascent stage. Therefore, the outcome of the study benefits the policy framers, researchers, and online course developers in providing valuable insights that affect behavior. Furthermore, the study forms a base for policy recommendations to aid decision-makers in designing effective initiatives that consider current/similar issues in online learning platforms. Learners’ attitudes, self-efficacy, and experience with online learning platforms can be enhanced by increasing the perception of usefulness and satisfaction. The outcome of the study aids instructors and academic institutions in designing the mechanism that explores the factors that make differences in learner engagement levels.

Universities and organizations are expected to benefit from this study to find an effective approach for learner engagement in an academic setting. It is recommended that decision-makers and practitioners contemplate proposed factors for formulating strategies for enriching engagement levels. The study will suggest that organizations consider the model in analyzing employee learning and development through an online learning platform; therefore, understanding the level of engagement would enable a better understanding of learners’ expectations. The findings of the study propose new insights and implications for the instructors, online learning platform managers, and learners; If the online learning platforms provide courses and services that make learners perceive the value of pursuing the courses, this assists the learners to deeply engage in the platforms. Therefore, the significant influence of the constructs examined in the study aid in shaping strategies and approaches those marketers and instructors need to adopt to enhance learner engagement, resulting in higher development and implementation success rate.

4.2. Limitations and future scope of research

To begin with, the study’s questionnaire was not entirely free of subjectivity as the questionnaire was released at a single point of time. Hence, the accuracy of the results might reduce as the user reactions tend to change over time and other influential factors. Further, studies should be conducted that are longitudinal in nature. Secondly, the study was limited to the users of online learning platforms where the perceptions of non-users were not captured. Future research should focus on a wider scope of the respondents for the generalizability of the results. Thirdly, the moderating effect of age, experience, gender, and voluntariness may impact the engagement levels as it may also account for good points for future research. Future studies can examine the underlying mechanisms for course quality, instructor quality, course difficulty level, and system-related factors on learner engagement. Besides, the sample for the study was drawn from the specific geographical location India, which effects the generalizability of the study findings. This warrants the need for a wider study by collecting data from diverse geographical locations to enhance the generalizability of the findings. Further, the study is limited to the current e-learning situation which is not supported by the artificial intelligence or advanced ChatGPT systems. The immense growth of technology has been a crucial contributing factor to the resurgence of artificial intelligence in education (Andrews et al., 2021). Hence, the advancement of AI on e-learning platforms will impact the learning styles, enable the learners to be more engaged, persist, and enhance their learning experience. Therefore, researchers have further scope of improving UTAUT framework to analyse the adoption of AI specific to e-learning. Finally, UTAUT is a widely adopted theoretical framework to examine and predict the individual’s technology acceptance and use behavior. In recent years, researchers have explored and examined the effectiveness of the UTAUT framework in several study contexts, especially in developing countries (Venkatesh, 2021). Furthermore, researchers have proposed modifications to the original UTAUT model in the last three years by adding additional constructs or modifying the relationships between existing factors to enhance its explanatory power and account for new variables. The factors of the UTAUT model and additional context-specific (such as cultural norms, socio-economic factors, educational policy, and level, etc.) factors have been reported to impact the individual’s behavior significantly (Venkatesh, 2021). This ongoing application indicates the enduring relevance and effectiveness of UTAUT as a theoretical framework. Although existing studies have provided valuable insights, there is dearth of literature examining the adoption intention and engagement behavior considering these context-specific factors in e-learning context. Therefore, further research is needed to validate and refine the UTAUT framework in specific contexts empirically.

5. Conclusion

The online learning platform is an innovative educational model for rendering education for learners. The existing body of literature has primarily emphasized behavioral aspects such as adoption and continuance intention. While there are still very limited scientific studies on assessing engagement and retention behavior. The current study is aimed at assessing the antecedents of the learner’s engagement behavior in the online platform at an individual level. Therefore, the conceptual model for the study was designed based on the UTAUT framework and scientifically validated by introducing learner self-efficacy as a predictor variable and e-learner engagement behavior as an outcome variable. The findings of the study affirm that flexibility, a favorable perception of utility, communal support, self-belief, and an uninterrupted supply of resources are the major means of enhancing the engagement levels of the learners. Therefore, changes among these attributes cause concurrent changes in the intention and engagement levels of the learners. The study contributes to the extant body of literature by empirically providing evidence for e-learner IUB as a full mediator between PE, EE, FC and SE and a partial mediator between SI and learner engagement. The findings of the study help the stakeholders such as educators and course developers to stand ahead in the market by paying attention to design strategies that increase the learner’s engagement levels in online learning platforms. To conclude, online learning is undoubtedly a source of competitive advantage. E-learner’s intention to use behavior has a significant impact on the e-learner engagement behavior, though there are direct influencing factors. Still the mediating role of IUB plays a vital role in enhancing the engagement of the learner, which in turn overcomes the low persistence levels.

Ethics statement

To ensure confidentiality, participants’ personal identifiers were removed prior to processing the data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Venisha Jenifer Dmello

Ms. Venisha Jenifer Dmello is pursuing Ph.D. at the Department of Commerce, MAHE Manipal. Her research interests include digital behavioral studies, consumer behavior and e-commerce.

Vadiraj Jagannathrao

Dr.Vadiraj Jagannathrao works as an associate professor with department of commerce, MAHE, Manipal. His research interests involve marketing, digital marketing, and customer engagement.

Ambigai Rajendran

Dr. Ambigai Rajendran works as an associate professor with department of commerce, MAHE, Manipal. Her research interests involve interdisciplinary research, digital behavioral studies, women and child studies.

Shilpa Badrinath Bidi

Ms. Shilpa Badrinath Bidi is pursuing Ph.D. at the Department of Commerce, MAHE. Her areas of interest consist of organization behavior, HRM, business analytics, behavioral studies, and application in higher education.

Tathagata Ghosh

Dr. Tathagata Ghosh is a Professor and Associate Dean (Academics) at TA Pai Management Institute, Manipal. His research interest includes marketing using the digital media, travel and tourism marketing.

Jaspreet Kaur

Dr. Jaspreet Kaur works at the Business School, Vivekananda Institute of Professional Studies. Her area of expertise is empirical research in social sciences, especially consumer behavior and sustainable purchase.

Kavitha Haldorai

Dr. Kavitha Haldorai is a researcher at the Dedman College of Hospitality, Florida State University, USA. Her research interests include HRM and organizational behavior in hospitality and tourism industry.

Appendices: Structural Model (Author’s own)

Figure A1. Path Model