Using machine learning to predict student retention from socio-demographic characteristics and app-based engagement metrics

Abstract

Student attrition poses a major challenge to academic institutions, funding bodies and students. With the rise of Big Data and predictive analytics, a growing body of work in higher education research has demonstrated the feasibility of predicting student dropout from readily available macro-level (e.g., socio-demographics or early performance metrics) and micro-level data (e.g., logins to learning management systems). Yet, the existing work has largely overlooked a critical meso-level element of student success known to drive retention: students’ experience at university and their social embeddedness within their cohort. In partnership with a mobile application that facilitates communication between students and universities, we collected both (1) institutional macro-level data and (2) behavioral micro and meso-level engagement data (e.g., the quantity and quality of interactions with university services and events as well as with other students) to predict dropout after the first semester. Analyzing the records of 50,095 students from four US universities and community colleges, we demonstrate that the combined macro and meso-level data can predict dropout with high levels of predictive performance (average AUC across linear and non-linear models = 78%; max AUC = 88%). Behavioral engagement variables representing students’ experience at university (e.g., network centrality, app engagement, event ratings) were found to add incremental predictive power beyond institutional variables (e.g., GPA or ethnicity). Finally, we highlight the generalizability of our results by showing that models trained on one university can predict retention at another university with reasonably high levels of predictive performance.

Effectiveness of Learning Management System of Universitas Negeri Malang in Supporting Distance Learning

“…This is in line with the theory of basic LMS activities, and that the existence of an LMS must facilitate learner registration, delivery and tracking of courses and learning content, testing, as well as enable the management of classes taught by educators (Cheawjindakarn et al, 2013). The first aspect that indicators the success of online distance learning is the online learning environment and how to remote learning environment (Matz et al, 2023). The online learning environment refers to where learners access online resources, use the system to access online courses and communication, get help from educators, and receive assessments.…”

Algorithmic Discrimination From the Perspective of Human Dignity

“…The possibilities of data aggregation, data reuse, data combination and inference, de-anonymisation and re-identification of individuals, categorisation, ranking, assessment, and individual or group profiling of individuals have greatly increased with AI (e.g., FRA, 2020;Smuha, 2021;Yeung, 2019). Some AI systems have been developed to make automated inferences about identity, personality-constituting traits, and other sensitive information such as emotions, character traits, mental states, or political orientations (e.g., Kosinski, 2021;Matz et al, 2023). AI-based biometric and psychometric evaluations (e.g., emotional AI) can be used for targeting (e.g., in marketing), risk assessment (e.g., in applicant selection or calculating the probability of dropping out of university or defaulting on a loan), and behavioural control (Valcke et al, 2021).…”

Predicting college retention rates from Google Street View images of campuses

“…From there, they used other algorithms to determine the model and make of these cars, their price, and their fuel range, and used publicly available information on political stances, race makeup, and percentage of people with high school degrees in each city, which offered a more complete approach to their research (15). Another study was successful in predicting student retention rates using other features, such as previous academic performance, socio-economic factors, or relationships in their family life, which provided a more holistic approach to understanding the factors that come into play when predicting retention rates that our model simply wasn't testing (16).…”