Augmented Reality, Physical Manipulatives, and Static Representations in Middle School Geometry: A Mixed-Methods Study of Spatial Skill Development

Abstract Middle school geometry requires students to coordinate two-dimensional views with three-dimensional structures. Although augmented reality (AR) may support this work, comparisons with static materials alone do not show whether its potential value extends beyond access to a manipulable model. This mixed-methods study examined a marker-based AR application alongside instruction using static representations and instruction using physical unit cubes. Seventy-four seventh-grade students from two public middle schools completed five lesson hours on drawing views of unit-cube structures. The contrasts were conducted in different schools and analyzed as two parallel, two-group quasi-experiments rather than a single three-arm study. Spatial visualization and mental rotation were assessed before and after instruction; eight students from the AR groups were interviewed about their experiences. Both AR groups improved on both outcomes, while both comparison groups improved on spatial visualization. After adjustment for pretest scores, no between-group difference was statistically significant. The estimated effects favored AR in all four comparisons, but the study lacked the precision to determine whether AR outperformed either comparator. Interviewees described using rotation to inspect parts obscured in static drawings. These findings suggest that AR can support spatial learning during ordinary geometry lessons, but do not establish its superiority over physical manipulatives or static-representation instruction. The study foregrounds the fit between representational affordances and spatial tasks while identifying the need for larger, better-controlled comparisons.