1. Introduction
Lv [
1] stated that in the recent decades, researchers and academics have paid more attention to machine learning as an instructional method. Particularly in the era of digitalization, machine learning is considered an alternative to the conventional learning model that makes use of the internet to deliver education in an unconventional manner and has become a strategic approach for educational institutes (Lallez [
2]). Mastan et al. [
3] documented that the machine learning approach that facilitates the accumulation of knowledge for each and every person plays a vital role in the accomplishment of educational institutions. Furthermore, Shahbazi and Byun [
4] highlighted that machine learning adoption can transform traditional methods of classroom instruction into digitalized instruction methods that can be used to enhance the performance of educational institutions. According to Almaiah et al. [
5], massive improvements in digital technologies change the educational environment, models, and practices, and particularly the learning paradigm. Therefore, educational entities that use various undated learning technologies to teach, learn, and interact with potential students are relevant for machine learning adoption, as suggested by Lallez [
2].
Lee and Lim [
6] stated in their study that advancements in technology exert pressure on educational entities to cope with these changes and adopt machine learning mechanisms for teaching and learning processes. Meanwhile, Hessen et al. [
7] highlighted the development of digitalization as instigating traditional ways of learning and teaching to undergo emerging changes. According to Almaiah et al. [
5], in the digital educational environment, IoT enables the integration of academic data, which can be used for the transformation of learning processes and flourishing e-learning activities. Zeadally et al. [
8] stated that in the era of digitalization, organizations mostly develop technologies such as the Internet of Things (IoT) in order to improve their capacity for improvising and learning in the future. Furthermore, educational institutes can digitalize their practices and improve e-learning activities by adopting emerging technology such as machine learning. The purpose of the current study is to highlight the role of IoT for the improvement of e-learning.
Ashima et al. [
9] documented that although organizations execute their instructional activities using various technological applications, deliberation on the role of IoT has been ignored. In the digital age, the multidirectional mechanisms of advanced technologies enable educational organizations to interact with learners, particularly students, in order to collaborate and participate in learning. Sinha et al. and Huang and Li [
10,
11] pointed out in their studies that in the recent scenario of a digitalized educational environment, organizations have put efforts into the procedures of IoT and building for the DEPs attached to these multi-directional IoT programs in their strategic educational planning. De Vas et al. [
12] highlighted the important of IoT and stated that IoT has gained wider attention as it has been considered a valuable strategic movement for educational entities, which increases the mechanism of digital educational platforms (Huang and Li [
11]). Therefore, the focus of the current study is to explore the role of IoT in the improvement of digital educational platforms.
On the other hand, Gillet et al. [
13] stated that DEPs improve the students’ and learners’ attachment to the learning processes of educational institutions, as well as enable the students to accept the changes made by the institutions in their learning processes. Liu and Ardakani [
14] documented that when students are highly attached to the institution regarding the execution of digital educational activities, they are more inclined to record higher machine learning adoption. Existing studies such as Almaiah et al. and Huang and Li [
5,
11] have highlighted that educational institutions with IoT provide massive opportunities for the involvement of students in their strategic learning decisions. According to findings of Sinha and Dhanalakshmi [
10], IoT raises DEPs; this ultimately improves machine learning adoption. The current study highlights the mediating role of DEPs in the relationship between IoT and machine learning adoption.
The current study adds to the existing body of knowledge by exploring the direct effect of IoT on the adoption of machine learning. So far, limited studies highlighted the role of IoT for the improvement of machine learning mechanism. As per the suggestions of Zeadally and Tsikerdekis [
8], IoT gains importance for the direction of machine learning. However, IoT is critical but not sufficient for the adoption of machine learning directly. Therefore, the current study contributes to the existing body of literature through intervening role of DEPs. A total of five parts make up the present study. The first part serves as an introduction, while the second discusses the context, the literature reviews, and the research methods.
Section 3 of this research paper presents and discusses the study’s findings. The study’s findings and commentary will be presented in
Section 4 and
Section 5, respectively.
3. Methodology
This was a cross-sectional study and questionnaire (See
Appendix A) was the main tool used for data collection, which was kept as concise as possible with simple/easy-to-understand wording. Cross-sectional studies can be generalized to the whole population from the basis of a small representative sample. It is a reliable and valid method with which to make conclusions regarding the issue being addressed in the study, paving the way for others to use it as a reference for future studies.
Random sampling technique was used for the selection of sample. The population consists of 91 public-sector universities. For the purpose of study, information about universities was taken from the Ministry of Education of China. In the next step, we approached the universities for the information about various departments. Out of 1678 departments, we selected 310 having IT infrastructure, students’ digital platforms, and e-learning infrastructure; most of these departments included information technology, management sciences, and public administration.
Table 1 depicts the demographic information of selected departments.
The chairman or head of department was surveyed using an online form built with Google Docs’s tools. Online survey tools were developed for this specific reason. We split the survey administration into two waves, T1 and T2, separated by a two-week time frame to avoid the potential for bias caused by the more conventional sampling approach. Selected respondents were sent a link to the online survey platform. For the purpose of facilitating a deeper comprehension of the concepts, the questionnaire was created in both English and Chinese. Between January 2022 and April 2022, email surveys were sent out to potential participants. T1 (the first wave) consisted of a survey being sent out to a random sample of 310 people. Information on the IoT and DEPs was gathered at TI. Only 312 valid replies were obtained in the first round. The second wave of data collection (T2) began after an interval of two weeks. Only 312 people participated in the second round of the survey. Department chairs were asked to judge how well their teams had embraced machine learning. At long last, we have 271 finished, fully-complete replies.
Measures
First section of the questionnaire was regarding demographics of the departments which included departments age, type, and size. Details are given in
Table 1 below.
The second part of the survey asked about experiences with the Internet of Things, data exchange protocol, and machine learning. There was a five-point Likert scale used for all the questions, with 1 indicating a strong disapproval, 2 a moderate one, 3 a neutral one, 4 an agreement, and 5 a strong agreement. To gather information from the designated university divisions, we used the following scales, which we either modified from existing instruments or created from scratch.
De Vass, et al. [
17] created a 10-item scale to assess IoT as an independent variable, and it was employed in this study. Moreover, mediating variable (DEPs) was measured with 8-item scale formulated by Rai and Tang [
30]. A 3-item-based construct of machine learning adoption was adapted from Jadhav [
31].
Table 2 shows details of constructs.
5. Discussion
The aim of the current study was to identify how IoT affects DEPs and to what extent DEPs help the adoption of machine learning in educational institutes. Our study consisted of four hypotheses, which explained the association between IoT, DEPs, and machine learning adoption. Study H1 explained the direct effect of IoT on the adoption of machine learning (0.17 **). The findings revealed that IoT is avaluable means of increasing opportunities for educational institutions to engage them and adopt machine learning. Arevalo-Lorido et al. [
32] suggested that technological development continues to rapidly grow, which changes the learning mechanisms of educational institutions, including machine learning adoption. IoT helps educational institutions in giving information which is computer-generated to all the stakeholders.
Study H2 shows the direct relationship between IoT and DEPs (0.29 **). The findings suggested that IoT provides opportunities to the organizations for the development of digital platforms. Previous findings, e.g., by Makkar and Kumar [
33], support that IoT is a significant predictor of e-learning activities of educational institutions. Waheed et al. [
25] documented that IoT facilitates organizations in building digital platforms and adopting machine learning. The prior findings of Nykyri et al. [
34] suggested that IoT is a set of all processes and applications of technological resources to develop platforms for acquisition and dissemination of data and information.
The third hypothesis links DEPs and machine learning adoption. DEPs act as important means for the adoption of machine learning. This study finding is constant with previous research, e.g., by Huang and Li [
11], about digital platforms through which organizations can easily adopt e-learning and machine learning. Fourth, DEPs mediate between IoT and the adoption of machine learning. Digital platforms generate the required information by using IoT and provide a foundation for the adoption of machine learning. It is impossible for machine learning to function without its fundamental building blocks, which are comprised of data sets, algorithms, evaluation, and output. Liu et al. [
14] documented that educational institutions try to build up DEPs by enabling IoT, which is apleasing and desirable factor for the adoption of machine learning. These findings revealed that DEPs in response to IoT positively influence the adoption of machine learning.
5.1. Theoretical Contribution
The work at hand considerably and conceptually adds to the current literature of information technology and e-learning. This work also considerably adds to the current body of information by increasing the UTAUT. The main contribution is the formulation of a model based on the facilitating conditions and social influence assumptions of UTAUT, which tested the IoT as a determinant of DEPs and machine learning adoption. There are limited research studies that consider the technological facilitating factors based on UTAUT for boosting machine learning adoption. Limited studies focused on the assumptions of UTAUT to highlight the determining factors for the adoption of machine learning. Furthermore, a comprehensive research model based on the assumptions of UTAUT was developed for educational institutes to test both the direct and indirect impact of IoT on machine learning adoption.
Furthermore, the current study also adds to the existing body of knowledge by explaining the role of IoT in building DEPs, which in turn enhances the e-learning approach of educational institutions. Huang and Li [
11] document that DEPs are an important mechanism and important for the thriving formulation of a new model for e-learning. Existing studies in the relevant field ignore the role of DEPs with respect to their determinants and outcomes. As a result, the current study took this into account, filled this research gap, and focused on IoT as a potential determinant of DEPs.
5.2. Practical Implications
The findings of the current study provide valuable implications for practice. First, the findings suggest that educational institutes can adopt the machine-learning method with the help of the IoT and through DEPs. When these institutions respond positively to digitization, adoption of machine learning is possible. Second, according to this study, educational institutions give attention towards the facilitating infrastructure, such as IoT, to make it possible to build digital platforms for the exchange of required information. The findings suggested that the managements of educational institution stake serious action for the development of infrastructure that supports the adoption of e-learning and machine learning procedures. Finally, the findings of the current study also suggested for the managements in practice to build educational digital platforms for the smooth functioning of learning activities. Such platforms provide a foundation for the exchange of valuable information required for the formulation and implementation of innovative e-learning techniques. Therefore, the managements can easily promote and adopt machine learning mechanisms in their institutes through proper IoT and DEPs.
5.3. Limitation and Future Recommendations
There are limitations to the current study that might be suggestions for future studies. Before moving on, it should be noted that the present study focused only on Chinese universities. However, in order to generalize the study’s conclusions, future research may expand its scope to include universities in other developed countries. Second, although we used a cross-sectional approach to gather data and evaluate our findings, qualitative studies may help us uncover other elements that might help push the needle on machine learning’s widespread adoption. Third, this study used UTAUT to examine the influence of IoT on DEPs and their adoption of ML; future research might benefit from exploring the role of alternative theories, such as a DOI and the TAM model.