Practical Adoption of Learning Analytics in Future Education
Learning analytics is the process of transforming raw student learning data into analytical insights, adding human intelligence to data using algorithms, and obtaining actionable insights to achieve the purpose of improving students' performance. During the process, researchers and practitioners will design observation measurements, conduct knowledge analysis, develop interventions, collect meaningful feedback and model learning outcome and retention.
With learning analytics, we can explore various types of questions! For students, how long did top 5\% students spend on watching videos, doing assignments, and online readings? For instructors, what course resources are students using? What course resources do student find the most satisfactory? For departments, how can learning analytics help predict how many students will want to register for a particular course? For advisors, which courses taken in first year are correlated with a higher GPA in second year?
Where could researchers get the learning data sources (input) from? Learning Management System (LMS). The most common data sources in LMS are 1) activity records which describes the users' interaction with the online platform and learning contents, 2) performance records which depicts the users' self-reported measures and results over the evaluations and tests, and 3) user profile information in which students' characteristics might impact the students' learning behaviours.
What are the main types of learning analytics? There are mainly descriptive and predictive analytics leveraged to meet the educational goals. Descriptive analytics provides insights about the past learning interactions and allows decision makings aimed at impacting future learning processes. Predictive analytics predicts the elements and variables that could impact ongoing learning processes and prompt educators to take proactive, preventative actions.
As my research focuses on asynchronous education and invention of machine-assisted educational toolkits, the subfields of learning analytics that I am actively participating in are:
- MOOCs
- Personalized dashboards and visualizations
- Participatory design
- Self-regulated learning
- Social learning
- Video analytics
- Intelligent tutoring systems
- Collaborative problem solving
- Personlized prompts and feedback
- Learning design
- Predictive analytics