UCLA

Personalized learning, which has the potential to raise student achievement, requires understanding the competencies of students. Visualizations can help provide thisunderstanding. Jeon et al. (2021) presented a latent space model that creates interaction maps visualizing response patterns from item response data. My dissertation proposes extending this latent space model to yield profiles that can visualize individual student competencies and developing a practical tool based on those profiles that can help promote personalized learning. Five aims will be achieved within the proposed study. First, I will investigate different formulations of the model that can yield different kinds of profiles. Second, I will introduce these profiles created from the interaction maps and the information they can convey. Third, I will investigate the validity and reliability of these profiles. Fourth, I will present empirical applications of this approach based on real-life datasets and validate this approach. Finally, I will present a web-based application that can create interaction maps and profiles from uploaded datasets.