A MULTI-INSTITUTIONAL STUDY OF LAB ACTIVITIES USING THE AUGMENTED REALITY SANDBOX: IMPACTS ON LEARNING

From Fall 2016 to Spring 2017, we developed and tested three different applications of an Augmented Reality (AR) Sandbox to teach introductory geoscience students about reading topographic maps at four institutions, including primarily undergraduate, master's granting, and R1 universities. These applications are characterized generally as 1) unstructured play, 2) a semi-structured lesson, and 3) a structured lesson. Students participated in one of the three conditions, and completed Topographic Maps Assessment items (TMA; Newcombe et al., 2015) and a series of mental rotation spatial questions a week later.

Previous studies suggest that there is a need to test different strategies using the AR Sandbox in order to produce learning gains and stimulate interest (Ryker et al, 2015; Woods et al, 2016). Simply asking students to “play” with the sandbox increases student interest, but not topographic map skill (Ryker et al, 2016). Similarly, walking students through a highly structured, instructor-guided exercise generates enthusiasm, but no significant gains in topographic map reading skills (Giorgis et al., 2016). However, students do perceive that they have learned more after using the AR Sandbox (Woods et al, 2016).

We compared application groups' scores on the TMA post test and mental rotation tests after completing the activity with the sandbox.There is no significant difference in either score between applications (structured, unstructured, semi-structured, or control), indicating that students in each application had comparable mental rotation skills, and that no one treatment worked universally better than another. Regression analysis shows that multiple factors significantly predict performance on the TMA including institution, spatial performance, and comfort before and after with topo maps. Prior courses, major, or application did not predict TMA performance. There is a significant difference in TMA scores between universities. Finally, some preliminary data suggests that students with low mental rotation abilities do better with a semi-structured than a structured assignment.

This presentation will discuss more detailed results from our analysis, lessons learned about teaching with the AR Sandbox from a pedagogical perspective, and needed future directions for research with this tool.