Transitioning From Studying Examples to Solving Problems: Effects of Self-Explanation Prompts and Fading Worked-Out Steps.

TLDR

Worked‑out examples, which include problem formulation, solution steps, and final answers, are essential for early skill acquisition in structured domains, yet fading these steps reliably improves near‑transfer performance but not far‑transfer. The study aims to improve far‑transfer by pairing fading of worked‑out steps with prompts that elicit identification of underlying principles. The authors implemented a fading schedule of worked‑out solution steps while introducing prompts that ask learners to articulate the principle illustrated by each step. Two experiments showed that this combined fading‑plus‑prompting approach yields medium to large gains in both near and far transfer without extending learning time, making it a practical and effective instructional strategy.

Abstract

Although research has demonstrated that successively fading or successively removing more and more worked-out solution steps as learners transition from relying on examples to independent problem solving reliably fosters performance on near-transfer tasks—relative to example–problem pairs—this effect is not reliable on far-transfer tasks. To address this, the authors combined fading with the introduction of prompts designed to encourage learners to identify the underlying principle illustrated in each worked-out solution step. Across 2 experiments, this combination produced medium to large effects on near and far transfer without requiring additional time on task. Thus, the instructional procedure is highly recommendable because it (a) is relatively straightforward to implement, (b) does not prolong learning time, and (c) fosters both near- and far-transfer performance. Worked-out examples typically consist of a problem formulation, solution steps, and the final answer itself. Research indicates that exposure to worked-out examples is critical when learners are in the initial stages of learning a new cognitive skill in wellstructured domains such as mathematics, physics, and computer programming (Anderson, Fincham, & Douglass, 1997). Moreover, studies performed by Sweller and his colleagues (e.g., Sweller &