K.T. Yuwono, H.D. Sujono
The aims of this study were to: (1) test the effect size of the research findings under the issue of the effects of e-learning on the students' learning outcomes, (2) investigate the combined effect sizes by doing the weighting of each finding, and (3) know the homogeneity of the research findings. A meta-analysis was conducted to synthesize research on the implementation of e-learning from Year 2000 to 2013. The search of the researh articles was done through several data sources namely ScienceDirect, Springer, IEEE, ProQuest, Sage, and BMC. The search was also conducted in libraries. In this study, there were 15 research articles selected based on the several predetermined criteria. The results of this study indicate that there is research with a negative effect size (-0.05), meaning that the learning outcomes in the implementation of e-learning are lower than those in conventional learning. There is a study with an effect size value of 0.00. Based on the value, there is no significant difference on learning outcomes in e-learning and conventional learning. The other studies have shown positive results; it means that the outcomes of e-learning are better than those of the conventional learning. However, there are 4 studies whose results are not significant. The combined effect size is 0.67 with a confidence interval of 0.42 to 0.91. Regarding this, the overall learning outcomes in the implementation of e-learning are higher than those in conventional learning. The result of the heterogeneity test using I2 formula is 71.95%, which means that the sample of this study is not homogeneous. Therefore, the generalization is not possible to be achieved. © Published under licence by IOP Publishing Ltd.
Faculty of Engineering, Yogyakarta State University, Indonesia
