Robust effects of the efficacy of explicit failure-driven scaffolding in problem-solving prior to instruction: A replication and extension. (October 2021)

• Record Type: Journal Article
• Authors: Sinha, Tanmay; Kapur, Manu
• Is Part Of: Learning and instruction. Volume 75(2021)
• Publication Date: 2021-10
• Subjects: Data science education -- Higher education -- Failure -- Scaffolding -- Problem-based learning; Learning -- Periodicals; Teaching -- Periodicals; Apprentissage -- Périodiques; Enseignement -- Périodiques; Learning; Teaching; Periodicals; Electronic journals; 370.1
• DOI: 10.1016/j.learninstruc.2021.101488
• Languages: English
• ISSNs: 0959-4752

Abstract

Abstract: Although Productive Failure has shown to be effective (Kapur, 2016; Loibl, Roll, & Rummel, 2017), it is not clear if failure in problem-solving is necessary. Initial work in a quasi-experimental setting suggests that explicitly designing for experiences of failure leads to better learning outcomes than designing for success. We build on this to report on a controlled experimental study where students are exposed to failure-driven, success-driven, or no explicit scaffolding in problem-solving prior to instruction. For assessments of non-isomorphic conceptual understanding, our results align with those from prior work. Despite the similarity in posttest scores, students exposed to failure-driven scaffolding demonstrate higher quality of constructive reasoning than those receiving success-driven scaffolding. Additionally, our study reveals learning benefits of failure-driven scaffolding (for both posttest scores and reasoning quality) on assessments of transfer. Several cognitive, affective and meta-cognitive mechanisms are investigated to explain robust learning benefits of failure-driven scaffolding in preparatory problem-solving. Highlights: Prior research has implicated that deliberate, guided failure may foster learning. Experimental work on explicitly designing failure-driven experiences is lacking. Failure-driven scaffolding in preparatory problem-solving facilitates deep learning . Deep learning stems from knowledge gap awareness and curiosity to fill these gaps. Negative affect, dissonance and metacognitive calibration also drive deep learning.