Re-Defining, Analyzing and Predicting Persistence Using Student Events in Online Learning

Abstract

In education, several studies have tried to track student persistence (i.e., students’ ability to keep on working on the assigned tasks) using different definitions and self-reported data. However, self-reported metrics may be limited, and currently, online courses allow collecting many low-level events to analyze student behaviors based on logs and using learning analytics. These analyses can be used to provide personalized and adaptative feedback in Smart Learning Environments. In this line, this work proposes the analysis and measurement of two types of persistence based on students’ interactions in online courses: (1) local persistence (based on the attempts used to solve an exercise when the student answers it incorrectly), and (2) global persistence (based on overall course activity/completion). Results show that there are different students’ profiles based on local persistence, although medium local persistence stands out. Moreover, local persistence is highly affected by course context and it can vary throughout the course. Furthermore, local persistence does not necessarily relate to global persistence or engagement with videos, although it is related to students’ average grade. Finally, predictive analysis shows that local persistence is not a strong predictor of global persistence and performance, although it can add some value to the predictive models.

A Comparative Study of Machine Learning Techniques for College Student Success Prediction

“…However, they do not discount the predictability of classical prediction models such as logistic regression. Similarly, a study by Moreno-Marcos et al (2020) in the domain of online education suggests that prediction accuracy is best achieved by random forest followed closely by logistic regression, both achieving higher than eighty percent predictive accuracy. In supervised learning, classical machine learning models such as logistic regression and newer used models, especially random forest, enable greater exploration in advanced learning environments such as online modality.…”

Understanding Utility and Privacy of Demographic Data in Education Technology by Causal Analysis and Adversarial-Censoring

“…Persistence, which captures the extent to which a student continues an activity for a long period, is another long-studied characteristic of students [66]. Students' persistence has been measured in different ways: Whitehill et al categorized students as persistent who interacted with the course at least once a week [92], while Crues et al identified three levels of persistence (low, medium, and high) based on the number of weeks students worked in the course [20].…”

Smart Learning

“…Sáiz-Manzanares et al [27] introduce a Moodle plug-in, so-called eOrientation, to detect at-risk students, which is used in junction with a learning analytics module, through which both supervised and unsupervised machine learning techniques can be applied. Moreno-Marcos et al [28] propose analyzing and measuring two types of persistence based on students' interactions in online courses: (1) local persistence (based on the attempts used to solve an exercise when the student answers it incorrectly), and (2) global persistence (based on overall course activity/completion). Chaparro-Peláez et al [29] details the architecture design, configuration, and use of the Moodle Workshop Data Extractor application, and proposes an initial validation of the tool based on the current peer assessment practices of a group of learning analytics experts.…”