Drawing from the scholarly literature, it seems that across all levels of education, teachers frequently must assume that students come to them without knowledge of statistics; in essence, instructors start from scratch each time they set out to teach statistics. As such, augmented reality (AR) presents an opportunity to explore ways for students to learn basic statistical reasoning skills authentically and engagingly. This study investigated the impact of an AR learning environment, MantarayAR, on college students' learning outcomes and the perception of collaboration and engagement. The design of this study was a 3 × 2 factorial design that compared and analyzed the performance of students given three types of augmented reality conditions; a high AR experience (with movement throughout a physical space), a low AR experience (without movement throughout a physical space), and no AR experience (2-D images and text), as well as two levels of collaboration (pairs and no pairs). Using the pre-test to categorize students into low or high prior knowledge, students with low prior knowledge assigned to either high or low AR experience had statistically significant higher learning gains than those assigned to a no AR experience. The results suggest that an AR experience was particularly advantageous for students without prior knowledge of statistical reasoning. Also, students in the AR conditions reported a higher perception of engagement as measured than those who did not.