ScienceSpace: virtual realities for learning complex and abstract scientific concepts

TLDR

The study builds three immersive virtual worlds to investigate how multisensory computer‑generated experiences affect learning of abstract scientific concepts, with future work including collaborative learning, intelligent agents, and comparison to 2D microworlds. The worlds target high‑school and early‑college students and cover Newtonian mechanics, electrostatics, and molecular structure and dynamics, with data collected via questionnaires, pre‑ and post‑tests, in‑situ predictions, experiments, and post‑session interviews. Results are inconclusive but suggest that students can improve mastery of abstract concepts in these virtual environments, and usability studies revealed problems that were addressed in subsequent refinements.

Abstract

Three virtual worlds have been built to investigate the effect of immersive, multisensory computer-generated experiences on learning topics in science. Currently targeted at high school and beginning college students, these worlds address Newtonian mechanics, electrostatics, and molecular structure and dynamics. Data has been collected on usability and learning through questionnaires, pre- and post-tests, in situ prediction and experiment and post-session interviews. The results are not uniformly conclusive but suggest that students can improve their mastery of abstract concepts through the use of virtual environments that have been designed for learning. Moreover, usability studies have identified many significant problems that have been addressed in successive refinements of these worlds. Future work will include collaborative learning studies (both local and distant), use of intelligent agents, and comparison with two-dimensional microworlds.