Multi_dimensional Quality Evaluation Modelling of E-learning Courses from Learner's Perspective

With the growth of e-learning in recent decades, research on the evaluation of e-learning standards sought to optimize and make e-learning effective. This study was conducted on the necessity of quality evaluation of e-learning systems from the learner's perspective, having the purpose of multi-dimensional quality evaluation modelling of e-learning courses using learner-based approach.
With respect to nature and purpose, the present study is an applied research, and with respect to the data collection method, it was a descriptive and survey type of research. The population of the study comprised all the students pursuing an MA in e-learning at Tehran University in 2013–14. Samples including 201 [F1] students were selected using a simple random method in 2014.
Data was collected using a researcher-made questionnaire to analyse the factors that influence the satisfaction of learners in e-learning courses. The questionnaire was validated by three professors of Education Sciences, while stability achieved using the Chronbach's alpha method (a=81%). To analyse data the path analysis model was used. The dominant influential components of learners’ satisfaction were introduced, and a causal model was designed. The influential factors for learners’ satisfaction and the factor's impact on the dependent variables were analysed. Finally, the suggestive fitting model was measured using Lisrel software version 8.5.
According to the obtained fitting indexes for the components, the indexes were very appropriate. The relationship between variables in the model of learner's satisfaction represented the significance of the relationship between the factor (learner's satisfaction) and latent variables (first-grade factors) at the level of 0.01%.
The results showed that, based on standard scores related to the variables that influence learner's satisfaction, it is possible to estimate learners’ satisfaction with e-learning using a good-fitting model and real-world data.