Is flipping effective? A meta-analysis of the effect of flipped instruction on K-12 students' academic achievement

Currently, the research on the overall effect of flipped instruction on K-12 students' academic achievement has been insufficient, especially across various subject areas and grade levels. To redress this research gap, this study first synthesised five aspects of research on flipped instruction, including theoretical underpinnings, application across disciplines, implementation benefits of FI, challenges and cautions in implementing FI, and guidelines for effective implementation. The study then reported on a robust variance estimation meta-analysis based on 53 effects from 27 empirical studies. A weighted average effect of d* = .54[(SE = .10),p < .001] indicated that flipped instruction promoted students' academic achievement moderately better than traditional classroom instruction. However, publication bias was revealed from the funnel plot, trim-and-fill, and the Egger's regression test. Given the large unconditional heterogeneity (I-2 = 81.2%), the mixed-effects moderator analysis shows that the publication type (published studies), grade level (middle and high school level), subject area, and study quality significantly moderate the effect size. Only the variable publication type explains a significant degree of effect-size heterogeneity. The implications for better implementing flipped instruction among teachers, school administrators, and policy-makers are discussed.