Does working memory moderate the effect of fading on math performance?

Background: Cognitive load theory suggests that novice learners may have limited working memory capacity; thus, difficult materials may impose cognitive load and prevent learning. To remedy this, proponents have suggested using various forms of worked examples to reduce cognitive load, such as fading or sequentially removing steps of a problem until the learner is solving it on their own. However, it's unclear how working memory may moderate the effect of these instructional strategies.

Aims: This study examined whether working memory influences performance on faded worked examples.

Methods: In the current study, 6th-grade participants (N = 114) were randomly assigned to (1) worked examples with self-explanation prompts, (2) fading with self-explanation prompts, (3) fading, and finally (4) problem-solving only conditions and assigned three links of homework (pretest, 3 days of the intervention, posttest) on geometry problems. An online computer tutor, ASSISTments, administered assignments.

Results: Results indicated significant differences in posttest performance amongst conditions, with fading demonstrating the largest effect sizes from pre to posttest. Further, only prior knowledge moderated the effect of fading.

Conclusions: Findings regarding theory and practical implications are discussed, such as specific recommendations for designing materials with working memory in mind.