The determinants of Saudi Arabians’ attitudes of E-learning System using GMDH and dce-GMDH models

Abstract

E-learning systems are considered one of the most important sustainable activities that always need research and development. Therefore, one of the most important challenges in the development of e-learning systems is determining the users’ attitudes. Many countries are working hard for improving the performance of their education systems particularly during COVID-19 pandemic. In this study, we examined the determinants of Saudi Arabians’ attitudes (SAA) of e-learning system. A well-structured questionnaire was designed to cover all aspects of our objectives. To answer the study's questions and test its hypotheses, a descriptive analytical technique was employed on 309 respondents. We found that SAA of e-learning system are good. We employed the Group Method of Data Handling (GMDH) technique, and the diversified classifier ensemble based on GMDH (dce-GMDH) to predict the SAA. The suggested approach considers Perceived ease of use (PEU), Speed and flexibility of work (FLX), Perceived usefulness (PU), Resistance to technology (RES), and Cost of technology (CT) as input variables whereas SAA as output variables. Several statistical tests are used to assess the efficacy of input variables in the collected data. The reliability and validity of input and output variables are acceptable because Cronbach's alpha (α) is greater than 70%, composite reliability (CR*) is greater than critical point 70%, and average variance extracted (AVE) is larger than 50%. Furthermore, there is no multicollinearity for the input variables because Variance Inflation Factor (VIF) is less than 10 and tolerance is more than 20%. The PEU, FLX, PU, and RES are found to have a significant positive impact on SAA under the ordinary least squares (OLS) technique. Additionally, the CT has a significant negative impact on SAA. Based on the error criteria, the findings show that dce-GMDH outperforms GMDH in predicting SAA

Keywords: E-learning, Artificial neural networks, Group method of data handling

JEL Classification: I2