Examining IoT-based smart campus adoption model: an investigation using two-stage analysis comprising structural equation modelling and artificial neural network

• Main Authors: Sneesl, Radhwan, Jusoh, Yusmadi Yah, A. Jabar, Marzanah, Abdullah, Salfarina
• Format: Article
• Language: English
• Published: Institute of Electrical and Electronics Engineers 2023
• Online Access: http://psasir.upm.edu.my/id/eprint/107811/1/Examining_IoT-Based_Smart_Campus_Adoption_Model_An_Investigation_Using_Two-Stage_Analysis_Comprising_Structural_Equation_Modelling_and_Artificial_Neural_Network.pdf; http://psasir.upm.edu.my/id/eprint/107811/; https://ieeexplore.ieee.org/document/10311591/

The progress and evolution of technology have been rapidly transforming various aspects of our society and daily lives, including colleges and campuses into smarter environments compared to the past. Despite the numerous advantages offered by cutting-edge technologies, such as IoT-based smart campuses, academic research on their implementation suffers from a significant lack of comprehensive information necessary to deliver efficient smart campus solutions. Therefore, the focus of this study is to investigate the significance of IoT-based smart campus adoption from 14 proposed hypotheses. The researchers collected data from stakeholders affiliated with universities in Iraq, resulting in a dataset of 442 observations. To analyze the data, a two-stage approach was employed, consisting of structural equation modeling (SEM) and reevaluated with the artificial neural networks (ANN) method. The findings provide evidence supporting the significance of various constructs. In particular, the model demonstrates satisfactory predictive relevance, indicating its effectiveness in making accurate predictions or forecasts. The ANN analysis suggests that the model has predictive capabilities. Moreover, the study findings support the importance of perceived usefulness in technology-specific factors, facilitating conditions, and propagation in organizational-specific factors, government support, social influence, and external pressure in environmental-specific factors, as well as privacy concerns, self-efficacy, satisfaction, and domain-specific knowledge in end-user-specific factors. Four hypotheses related to perceived ease of use, service collaboration, habit, and innovativeness were rejected. Notably, the study identifies propagation as the most significant predictor in the ANN analysis. The conclusions of this study can be beneficial for university administrators, manufacturers, and policymakers in understanding the essential components of smart campuses to enhance the adoption and maximize the effectiveness of smart solutions.