For a Learning Analytics class in Spring 2024, I completed a detailed proposal for a learning analytics platform. This project demonstrates understanding of text mining, LIWC-22, and self-regulated learning theory. Here are a few key details.
Problem
A positive correlation between vocabulary knowledge and academic success is well-supported in literature and research. Academic vocabulary deficits in writing are a stumbling block to students beginning college. Can we improve early college success via academic vocabulary? Additionally, students’ vocabulary knowledge is frequently assessed using in-class exercises and tests. Ultimately, learners need to transfer this knowledge to their working vocabularies to maintain the gain in the future. Can the ability to track learners’ digital writing paired with our increasing reliance on digital communication more accurately judge learners’ knowledge of vocabulary?
Goal
Enable students to monitor their development of a productive academic vocabulary to improve college outcomes.
Method
We would build a data store and text analysis mechanism. Instructors would select which words from the Productive Academic Vocabulary List (PAVL) they have taught. Students’ academic writing and social media posts would be fed into the system. Researchers would use text mining to detect and count students’ organic uses of the vocabulary words across sources. By incorporating casual and formal contexts, we could judge both near and far transfer.
Natural language processing would analyze the context to determine whether the term was used correctly.
The learning analytics tool would produce a customized dashboard for each student. Their dashboard would show their personal vocabulary acquisition statistics, including links back to the source material. Use will be broken out between near and far contexts. The dashboard is designed around Self-Regulated Learning theory to empower the individual to control their own experience.
Presentation and Report
For more information, view the full presentation and report below.