Cyberlearning: Detecting and Predicting Procrastination in Online and Social Learning

点击翻译按钮获取中文摘要

As online education becomes increasingly available and trusted by both employers and students, many workers are turning to online courses to advance their education and job prospects. However, online courses demand effective time management skills, as students are required to plan and set goals, manage their time, and work by themselves (or in a group), often with less structure than an in-person course. This increases the risks of procrastination, a key challenge to time management and success in both work and education contexts. To address those risks, this project will use computational algorithms to model students' procrastination behaviors, identify indicators of likely future procrastination, and detect it early on in both individual and in group work. The algorithms will learn to predict procrastination according to learners' studying behavior captured by a time management application and their performance in courses. The findings of this project can be used to enhance students' learning by helping them to set goals and plan their work, monitor their progress, and keep track of what they need to do to successfully accomplish their assignments on time. These findings can be applied to related areas such as workforce development, and the data collection tools and algorithms developed will be made available to other researchers who want to work on related questions at the intersection of behavior and learning.This project examines individual and group procrastination behavior by developing computational models using data on students' self-reported cognitive, metacognitive, motivational, and affective processes. Current theories of procrastination will be studied and extended based on cross-sectional self-report survey data asking for student self-ratings of procrastination related to academic tasks, and time-stamped trace data of studying and interaction behavior generated by a mobile app used by students during their courses. The cyberlearning advancements of this study are (1) a novel model of individual and individual-in-group (social) procrastination, to detect procrastination based on both self-report and trace data; (2) a novel model to predict student performance based on their procrastination, previous task accomplishment behavior, and previous performance; and (3) exploration of the most parsimonious combination of self-report and trace data to produce effective procrastination model. These goals will be accomplished by (a) developing and updating an application for data collection and survey administration, (b) deploying the app in several graduate online courses, (c) analyzing data to understand underlying procrastination processes, and (d) developing machine learning algorithms to model and detect procrastination. The project will result in the dissemination of findings and developed algorithms to the broader field of sequential data science.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.