Date

2021-9-7

Author

Yürüm, Ozan Raşit

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

464
views

415
downloads

Cite This

Video-based learning has become very popular together with the increase in the number of massive open online courses (MOOCs). Learners interacting with video naturally exhibit viewing behaviors based on clickstream interactions such as pause, forward, and backward. These are important inputs to predict learning variables and to develop an intervention framework. However, there is no guideline or method on how to use these behavioral patterns in favor of learners. The purpose of this study is to investigate learners’ video clickstream behaviors, and to use them to develop a framework in order to increase the impact of video lectures. Accordingly, three experiments were performed. In the first one, a pilot study with 22 students was conducted. In the second one, a more comprehensive experiment was performed with 29 students to develop the framework. In the third one, an experiment based on one group pretest-posttest design was conducted with 18 students to evaluate the framework. These studies reveal important insights in terms of both the framework and comprehensive learning analytics. The results for the framework show that students’ satisfaction increases significantly in interactive videos created based on the framework when motivation is controlled. In addition, students’ frequency to go back to important points decreases significantly and students’ frequency to skip unimportant points increases significantly in interactive videos. Apart from the framework, the results show that the mean of backward speeds, number of pauses, and number/percentage of backwards are the most important indicators in predicting students’ grades. Furthermore, the study also shows that video clickstream behaviors are able to predict grade whereas the reasons behind the behaviors are also necessary to have to be able to predict satisfaction successfully. However, video clickstream behaviors and their reasons are not enough to predict engagement significantly.

Subject Keywords

Intervention Framework, Learning Analytics, Educational Data Mining, Video-Based Learning, Interactive Videos

URI

https://hdl.handle.net/11511/93046

Collections

Graduate School of Informatics, Thesis