From Visualization to Collaborative Inquiry: A Comparative Study of Markerless Augmented Reality Support of Scientific Self-Efficacy and Collaboration in Chemistry Learning

Abstract Chemistry inquiry experiments play an important role in developing students’ evidence-based reasoning, scientific practices, and model-based understanding. However, classic experiments such as Lavoisier’s 12-day bell jar experiment with mercury and red phosphorus combustion are difficult to conduct in authentic classrooms because of safety concerns and practical constraints. To address this issue, this study designed a markerless augmented reality (AR)-supported collaborative inquiry activity on the topic of the “composition of air.” A quasi-experimental design compared the AR-supported activity with video-based instruction among 222 grade 9 students from four classes. Pretest and posttest records were matched separately for the knowledge test (AR n = 71; video n = 84), the Chemistry Learning Self-Efficacy Scale (CLSE; AR n = 108; video n = 104), and the Collaborative Learning Experience Scale (CLCL; AR n = 90; video n = 110). Both groups achieved significant knowledge gains (Cohen’s dz = 1.52 and 1.67, respectively; ps <0.001), but no adjusted posttest difference was found between the AR and video groups (p = 0.769). Students in both conditions improved in total scientific self-efficacy; after Holm correction, the only significant adjusted between-group difference was in scientific communication, favoring AR (p = 0.026). No CLCL outcome differed significantly between groups after baseline adjustment and Holm correction. These findings suggest that both formats supported immediate knowledge learning, whereas evidence for an AR-specific benefit was limited to scientific communication self-efficacy.