E-learning experience: Modeling students’ e-learning interactions using log data

Abstract

This study aims to examine e-learning experiences of the learners by using learner system interaction metrics. In this context, an e-learning environment has been structured within the scope of a course. Learners interacted with learning activities and leave various traces when they interact with others, contents, and assessment tasks. Log data were formed on these e-learning interactions. In the data analysis phase, firstly, a data pre-processing was performed, and then confirmatory factor analysis (CFA) was used to test how well the measured learning activity variables represent the latent system component variables. Then it was tested whether these components compose a latent e-learning experience variable (second-order CFA). The results showed that the learners interacted with five different system components: hypertext, the content package, video, discussion, and eassessment. In conclusion, there is a factorial relationship between the system components and learning activities. These components taken together constitute an elearning experience variable. When the factor loadings between the e-learning experience structure and subcomponents were examined, the discussion interactions in which the learner structured knowledge highlighted. In summary, the discussions, formative assessments, and content activities formed the learners' e-learning experience together. In order to form a well-structured e-learning environment, these activities together should be experienced by the learners. Research Article * This article is based on PhD thesis titled "Determining user profiles for adaptive feedback system design" completed by the first author under the consultancy of the second author.

The effect of internet of things education through distance education on student success and motivation

“…As a result of this study, it was seen that Google Classroom, WhatsApp and Edpuzzle platforms used in the research process gave positive results in providing distance Internet of Things education. In this respect, it can be said that the study supports the views of Keskin and Yurdugül (2022) and Tianbo (2012). In addition, the current study showed that IoT concepts and technical skills can be acquired by students from different engineering programs that constitute the study group, and similar results were obtained in the related studies (Kleinschmidt, 2021;.…”

“…As Milman (2012) states low-quality and very long course videos can be a problem for students: thus, the results of this study are consistent with Milman (2012). Keskin and Yurdugül (2022) suggest that when designing e-learning courses, discussion forums should be included to ensure collaboration and communication between learners and e-assessment tools that students can test themselves. Many web platforms that provide online learning and discussion opportunities with graphical user interfaces significantly shorten the learning process and project development time of students (Tianbo, 2012).…”

Examination of the relationship between the attitudes of teacher candidates towards technology and their opinions of distance education

“…The first finding of this study is that the attitudes of teacher candidates towards technology are moderate, and that they have a positive evaluation of the distance education process. In the students' online learning interactions model, while the discussions, formative assessments, and content activities comprised the online learning experience of students, the main component is seen to be how discussion interaction facilitated the construction of student knowledge (Keskin & Yurdugül, 2022). Accordingly, in this research case in which the AYDEP platform was employed, the course process was enriched with discussion sessions, as well as interactive interactions, both in the lecturers and content of the course.…”

AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions

“…We recommend elicitation of learner's preferences using explicit and implicit data gathering techniques. Preferably, a diagnostic test should be presented to a learner [32]. This test can consist of content of different modalities and an assessment for each modality.…”

“…7) Feature Engineering of Learner-Content Interactions The data of learner-content interactions in its basic form does not reveal patterns. We need to perform feature engineering on the raw interactions data to come up with meaningful patterns [32]. To model the preferred modality detection as a machine learning problem, there is a need to extract learner's behavioural signals from the data.…”