Keywords

1 Introduction

The education sector, especially teaching and learning, has continued to experience rapid changes due to evolving technologies. There is a clear paradigm shift from the traditional teaching method to a more technological one. In the traditional system, learners were more passive, and the educator was considered the transmitter of knowledge. In the new order, through the availability and use of technologies, students/learners become more active participants and consumers of knowledge. The internet has made education more possible and democratized through effective online education [1]. Nonetheless, issues like non-availability and reach of technology may inhibit e-learning [2, 3]. The opportunities in online education necessitated by COVID-19 pandemic need to be explored as they are expected to be standard practices in the future [4]. One such opportunity is virtualization. Virtualization is taking place in many contexts, including education (e-learning), shopping (e-commerce), and friendship development (social networking sites and virtual worlds) [4,5,6]. However, some processes are more amenable to virtualization than others. This is the case in e-learning. Hence study aimed to identify the factors that affected virtualization in e-learning among university students in Pakistan through an online survey. Process Virtualization Theory (hereafter PVT) describes how amenable a process is to being conducted without physical interaction between people or between people and objects [1]. For this study, e-learning process virtualization referred to the vulnerability of a student–teacher virtual interaction since literature has suggested that the virtual learning outcome can be evaluated based on the attributes related to the learning ability of students [2]. Furthermore, Scholars [1] indicated that formal education might be virtualized under certain situations and for subjects. Students’ e-learning engagement in the virtualized learning environment can affect the student’s performance [4]. Virtual outcomes are therefore determined by the e-learning process virtualization in the virtual learning environment. Consistent with this idea, the current study investigated the e-learning process virtualization as virtual outcomes. Therefore, the study employed process virtualization theory as an analytical lens to explore the factors that restrict or facilitate the e-learning process virtualization. The four primary constructs of the process virtualization theory are sensory requirements, relationship requirements, synchronism requirements, and identification and control requirements that affect whether a process is amenable or resistant to being conducted virtually [1, 3]. In the current study, the two primary constructs, the sensory and relationship requirements, are considered since the other two constructs do not have much relevance in e-learning and can be considered in future studies. Moreover, the study considered how representation and digital divide as two moderating factors affected the virtualization of e-learning. The paper overviews the PVT theory focusing on the e-learning process virtualization delineated in the following section.

2 Literature Review

2.1 Process Virtualization Theory

One influential and comprehensive theory is the Process virtualization theory (PVT) to describe virtual processes. The PVT, proposed by Overby [1], has gained attention in several domains such as e-commerce [5], health [6], governance [7], e-learning [2, 4, 8], and information technology [9]. The strength of PVT lies in postulating interaction and the tradeoff between more pragmatic factors in developing one’s inclination towards the virtual process [5]. Secondly, the nature of constructs such as sensory requirements is very much applied in nature in several contexts [6]. Thirdly, based upon the above factors, PVT has parsimony and is suitable for explaining several attributes and expected relationships in different circumstances [7]. For example, PVT assumed that one’s sensory requirement is a function of one’s perception of the representation of the virtual process. There is an inclination to adopt a virtual process when that virtual process fits the practical requirement. Lastly, the PVT is a comparatively new theory; therefore, Overby [1] considered it timely in defining the phenomenon of virtualization. Although this theory has not been widely tested yet empirically [10], there have been many suggestions for its application in different settings. For instance, Graupner and Maedche [11] utilized PVT in the banking sector. Similarly, numerous scholars have enhanced it by turning it into a multiclausal mode for qualitative and quantitative research [7, 12]. Researchers found that virtualized processes have some limitations due to high sensory and identification requirements. Therefore, applying PVT in an e-learning process is a step in the right direction to gain new insights in an education context.

2.2 E-Learning Process Virtualizability

Virtualization theory advances the understanding of the willingness and participation of individuals in the virtual process [8]. In this regard, PVT proposed process virtualizability as the ability to perform a process without the involvement of any face-to-face interface among individuals or individuals and objects (e.g., computers). To this end, PVT outlines the critical factors in determining any virtual process’s outcomes [9]. Overdi (2008) operationalized virtualizability based on either (a) the quality of the computer-mediated interaction between stakeholders or (b) the adoption of the underscored virtual process. Process virtualizability is proposed to tap attributes, including the extent of the likelihood of the adoption or quality of the process in question. Previous research indicated that virtual processes related to shopping processes, such as electric commerce [5], e-government [7], and IT adoption in the industry [9, 13], are amenable to virtualization. Given that students’ e-learning engagement in the virtualized learning environment can affect the student’s performance, virtual outcomes are therefore determined by the e-learning process virtualization in the virtual learning environment [8]. Consistent with this idea, this study has employed process virtualization theory as an analytical lens to explore the factors that restrict or facilitate the e-learning process virtualization.

2.3 Sensory Requirements and E-Learning Process Virtualization

Sensory requirements are coined as one’s requirements and expectations from the virtual process regarding broad sensory experiences related to the process, other stakeholders, and objects [6]. The PVT suggests that when participants engage in a virtual process, they presume to sense several sensory involvements, including (1) seeing, (2) hearing, (3) smelling, (4) touching, and (5) identification of other stakeholders (e.g., instructor). These sensory requirements diminish the usage of virtual processes [14]. The overall physical experiences of an individual would shape the extent of their involvement in the virtual process in question. For instance, higher expectations of these sensory requirements can result in affective dissonance. Ample research has affirmed that individuals with higher sensory requirements could engage less in the virtual process [11, 15]. These studies clarified that sensory characteristic function as the troublesome antecedent of the virtual process outcome [14]. In the context of e-learning process virtualization, several research studies have evidence that learner-material interaction is essential for effective learning [8]. The rationale is that the lower sensory requirement (SR) of a process, the more suitable it is to be done virtually and vice versa. Similarly, learning materials’ audio and visual requirements could improve learning outcomes’ quality [4]. As a result, it would be challenging to execute the more demanding sensory needs and manipulate physical objects in the virtual environment. COVID-19 has forced all educational institutions to adopt a virtual learning mode. A greater need for sensory requirements reduces virtual outcomes at different educational institutions [10, 12]. Based on this, the following hypothesis was stated on the sensory requirements for e-learning process virtualization:

H1: Sensory Requirements have a negative influence on e-learning process virtualization.

2.4 Relationship Requirements and E-Learning Process Virtualization

Relationship requirements (RR) are necessary for process participants to interact with one another in a social or professional environment [15]. Overby [1] asserted that social interactions allow people to gain information, build trust, and form friendships. When everything else is equal, a process with a greater connection will be more resistant to virtualization [3]. As a result, the less virtualization-friendly the process is, the more participants favor RR. Prior research has indicated that RR negatively influences consumers’ inclination to adopt digital procedures [2, 10, 12]. Relationship requirement evidence proves that the strong bonding between students and teachers may affect learning outcomes. Previous research has shown that student–teacher bonding improved learning outcomes and academic performance at all levels of school [16]. The research identified that the connection between trust and intimacy is essential in a virtual context [1]. A virtual environment, on the other hand, makes it difficult to build trusting relationships. Since all brick-and-mortar learning is being replaced by virtual learning in COVID-19, students and teachers have had a bond before commencing virtual learning. However, in informal education during COVID-19, the connection between teachers and students had already been created. Therefore, we proposed the following hypothesis:

H2: Relationship Requirements have a negative influence on e-learning process virtualization.

2.5 Moderating Effects: Perceived Representation of ICT Orientation

Overby [1, 15] refers to the moderating constructs as the characteristics of the virtualization process. There are three moderating constructs: reach, representation, and monitoring capability [7]. Each moderating concept positively impacts the relationship between major concepts and process virtualizability. Even though IT plays a crucial part in virtualized processes, not all processes are virtualized using IT [3]. Furthermore, according to Overby [15], process virtualizability may be positively affected by the moderation of IT factors. The role of the moderating IT-based constructs, such as the ICT orientation of a system, has been overlooked in past studies [10, 11]. Therefore, currently, moderating components of the PVT are re-conceptualized in this study. ICT orientation of e-learning management system is relevant to the research setting. Virtual processes have expanded in recent years due to advancements in IT, notably the internet. Process virtualization is therefore influenced by IT [2]. A process virtualization’s initial IT component is representation, described as IT’s ability to show process-relevant information, including simulations of people and things in the real world, their attributes, and how we interact with them visually. ICT capability was used as a representation for the current study [10]. Examples include the sensory components involved with an interaction between students and learning material that may be simulated using IT. Sensory requirements can be integrated into a virtual process using IT. Computer simulations may be more effective learning tools in certain situations than physical objects [6]. Literature is evident that a higher degree of representation by incorporating ICT‐based features to address sensory requirements into virtual processes can enhance the chances of individuals engaging in the process [11]. Hence, building on the previous literature [11, 14], the processes with better representation capabilities through ICT would positively moderate the primary influence of sensory requirement on process virtualizability. Thus, we hypothesize that.

H3: The perceived representation (ICT orientation) of the learning management system moderates positive nexus between sensory requirements and e-learning process virtualization; thus, higher ICT orientation leads to higher e-learning outcomes.

Similarly, a steam of research advocated that the representation capabilities of e-learning management systems can address the requirement of students to develop a relationship by providing them opportunities to interact with instructors [8]. Some studies also stated that interaction between students and teachers is indispensable for students and a critical factor in determining their participation in e-learning undertakings [8] [18]. The nonexistence of opportunities for such interactions causes students discontent and affects their active participation during e-learning [4]. Considering the e-learning setting, students who lack interaction capability in the e-learning management system start perceiving complications in developing and upholding relationships with teachers. Conversely, modern systems with these representation capabilities, such as question and answer, live lecture streaming, feedback facility, etc., lessen the feelings mentioned above. Therefore, we assumed the following hypothesis:

H4: The learning management system’s perceived representation (ICT orientation) moderates the positive nexus between relationship requirements and e-learning process virtualization; thus, higher ICT orientation leads to higher e-learning outcomes.

2.6 Perceived Digital Divide

The PVT proposed a factor of reach and coined it as the capability of any virtual process that permits process participation [5]. For example, the internet connection availability and strength enable people to participate in a process. Thus, reach is a critical aspect while someone is undertaking the virtual process. Better reach can provide the flexibility to participate in a process (e.g., e-learning) anywhere in the world at any time [3]. To this point, reach is a key factor in determining process involvement across time and space. IT allows many educational processes to be conducted at any time and from various locations. IT facilitates a live, interactive session during virtual learning through Zoom, Google Meets, Google classroom, Second Life, and many other apps that allow people to interact, negotiate and give feedback in real-time. During synchronous virtual learning, IT-based virtual learning allows the connectivity between the dispersed students in real-time settings. Originally in PVT, reach moderates the relationship requirements and process virtualization linkage to determine the adoption and outcome of the process virtualizability [15]. For this study, the context of a developing country has been considered as reach is a critical issue, and digital divide prevails given poor connectivity and strength. The students residing in developing countries have greater concerns about digital divide. Even when they have to participate in online classes willingly yet, they perceive digital divide (e.g., internet issues) as a major difficulty in attaining online education and thus become less inclined toward e-learning Process Virtualization; we, therefore, hypothesize that:

H5: Digital divide negatively moderates the positive nexus between relationship requirements and e-learning process virtualization; thus, a higher Perceived Digital Divide (PDD) leads to lower e-learning outcomes.

3 Method

The study used a quantitative online survey for the data collection. The use of online data collection is suitable as the context of this study is the virtual process. An online questionnaire was distributed among the university students currently or previously enrolled and using a learning management system. The sample was approached using university volunteers, lecturers, and online platforms. A filter questionnaire asked, “have you ever attended e-learning/online classes” with a request to terminate the participation in the survey if they never used—the learning system. In total, 362 responses were obtained. Most of the participants were 51.7% female and 48.3%, male. Most participants were aged between 18 and 23 (64.9%); most students were enrolled in undergraduate programs 58.8%). The majority reported monthly incomes between 25,000 and 50,000 PKR (32.2%) for the family income.

3.1 Instruments

The current research adopted the items to represent the variables underpinned in this study from previous literature [1, 12, 15] with slight modifications. All items were measured on the five-point Likert scale anchoring “1 = Strongly disagree to 5 = Strongly agree”. Three items were used to measure each construct of Perceived Sensory Requirements (SR), Perceived Digital Divide (PDD), and Perceived Representation (PRICT). At the same time, four items were used to measure each construct of Perceived Relationship Requirements (RR) Process Virtualizability (PV).

4 Results and Analytical Strategy

The demographic analysis revealed that Descriptive analysis was carried out to prepare data for the SEM that involves; deletion of the outliers, normalization of data, and variance inflation (henceforth VIF) test. Consequently, 23 cases were removed from the total of 362 cases to normalize the data due to outlier characteristics of these cases. Pearson’s correlation matrix suggested a significant relationship between SR, RR, PDD, PRICT, and PV.

4.1 Confirmatory Factor Analysis

As mentioned earlier, SEM has been used to test the validity of the measurements involved in this study. Thus, the measurement model was initially tested for the goodness of fit test using the suggested six indices on AMOS. The results of CFA revealed suitable goodness of fit such as ×2/df =, CFI = 0.92, NFI, TLI = , RMSEA = and SRMR = 0.061, NFI = 0.97, and RMSEA = 0.036. One item from the construct of the PV was deleted to attain the model fitness. Later, validities assessments were carried out, the constructs revealed decent loadings and resulted in satisfactory composite reliability, and average variance extracted was achieved. Following the item loadings, Fornell-Larcker criterion-based (henceforth FLC) values were observed to assess the discriminant validity. The values of the FLC presented in Table 1 revealed the adequate and recommended parameters (CR = ≥0.7 and AVE = ≥0.5). Besides, the square root values of AVE were found to be less than the correlation between five constructs, PRICT, SR, RR, PDD, and PV, and thus, the assumption of the discriminant validity.

Table 1 Validity statistics
Full size table

4.2 Hypotheses Testing

We employed the CB-SEM method using AMOS software for the hypotheses analysis. Two hypotheses focused on the direct influence of sensory (H1) and relationship (H2) requirements on the e-learning PV, respectively. The results showed a direct and negative influence of the sensory requirement on the e-learning PV (β = −0.23), and H1 was supported. Likewise, results also verified that relationship requirement negatively influences—learning PV (β = −0.31), and H2 was supported. Overall, this model revealed a variance of 42%. Following the Preacher et al. [17] stage-wise moderation analysis approach, another model was run by adding three interaction terms using the bootstrapping method. The conceptual model hypothesized the three moderations, and all were added in this model by computing the interaction terms such as (SR-X-PRICT), (RR-X-PRICT), and (RR-X-PDD) to test H3, H4, and H5. The results indicated that PRICT positively moderates (SR-X-PRICT β = 0.36) the sensory requirements (H3) relationship with e-learning PV. The results suggested that PRICT positively moderates (RR-X-PRICT β = 0.19) the relationship requirements (H4) with the e-learning PV. Finally, the results indicated that PRICT inversely moderates (PDD-X-PRICT β = −0.26) the relationship requirements (H5) relationship with the e-learning PV. Model 2 extracted a variance of 57%, and a 15% moderate change in the variance remained moderate. Hence H3, H4, and H5 were supported.

5 Discussion

In the e-learning process virtualization context, there is evidence that learner-material interaction is essential for effective learning [8, 16]. The rationale is that the lower sensory requirement (SR) of a process, the more suitable it is to be done virtually and vice versa. This was the case in this study where there was poor student engagement in e-learning due to a lack of material interaction. Therefore, better ICT representations can improve students’ engagement in the virtual process of e-learning. More representation implies higher integration of multimedia facilities in ICT that can reduce or make up for the sensory requirements of the students (see Fig. 1). On relationship requirements, the findings of this study validate past research that indicated that RR has a detrimental influence on consumers’ inclination to adopt digital procedures [2, 12]. This particular result is a pointer that relationship requirement—evidenced in substantial bonding between students and teachers—improves learning outcomes and academic performance at all levels of school [18].

Fig. 1
A block diagram. The arrows from the blocks for perceived digital divide, perceived representation, sensory requirements, and relationship requirements lead to process virtualizability. The values on the arrows are minus 17, 64, minus 23, and minus 31, respectively.

Structural model

Full size image

Therefore, there is a need to increase the effort required to virtualize e-learning systems by providing apps that facilitate student–student, and student–teacher interactions, to diminish the participants’ belief that relationships can be attained through face-to-face only. It was evident from the result that digital divide serves as a negative factor as it decreases the probability of students’ engagement in a virtual learning environment, especially in the case of developing countries. This supports the earlier assertion that Internet reach allows virtual process participation [5], i.e., internet connection availability and strength enable the students to participate in virtual education anywhere in the world at any time [3]. Thus, efforts should not be spared in providing accessible, adequate, and affordable internet to all, especially in developing countries. This will make e-learning affordable and worthwhile in developing (e.g., Pakistan) compared to the nations with higher income levels (e.g., UAE). Thus, the students and other stakeholders would evaluate the perceived utility of an e-learning system (e.g., a learning management system) in terms of its capability to provide an alternative but relevant and practical learning facility. Consequently, we conclude that better representativeness of ICT (learning management system) through incorporating multimedia components, recorded lectures, interaction with instructors, availability of online materials and options to ask questions and get customized responses from instructors, submitting assignments, and getting customized comments), and eradication of digital divide can lead to the thriving e-learning environment.

6 Conclusion

COVID-19 pandemic has unexpectedly forced educational institutions to adopt virtual management systems to endure academic activities. Several countries face severe challenges due to the lack of internet access, digital infrastructure, and unproductive learning management systems. This research fills this gap by underpinning PVT theory to understand how the virtual process requirements of the e-learning process outline students’ involvement in effectively using the e-learning systems in place. This research provided a unique understanding of e-learning management systems to facilitate the educational institutions “going virtual.” Two significant findings of the study are that better representativeness of learning management systems and eradicating the digital divide can lead to a thriving e-learning environment. Therefore, countries like Pakistan and UAE must focus on delivering learner material and the e-learning systems’ attributes to ensure that the student’s sensory and relationship requirements are contented. Thereby, improving interactive capabilities that can address such necessities are significant components of attaining an efficacious e-learning process among middle east students.