Why working memory capacity changes who benefits from worked examples
Solving a math problem from scratch requires holding the goal, the intermediate numbers, and the rules for the next step in mind all at once. Working memory capacity is the amount of information a person can keep active and manipulate at the same time, and it is limited [2]. Worked examples change the task: instead of generating each step, the learner can study a completed solution and see how the steps fit together. That reduces the number of things that must be held in working memory simultaneously [1][3].
The moderation follows from that reduction. If a learner already has enough capacity to solve the problem without becoming overloaded, a worked example offers less additional relief. If capacity is low, the example removes a bottleneck that would otherwise interfere with learning the procedure. This is why the benefit is expected to be larger for lower-capacity learners, not because high-capacity learners are harmed, but because they have less room to improve from the same support [3].
What the classroom studies actually show
A study of faded worked examples in mathematics instruction examined how working memory capacity moderated the effect and found that capacity played a role in how much learners benefited [1]. The same pattern appears in a review of worked examples, prior knowledge, and cognitive load: learners with lower working memory capacity were found to derive more benefit from worked examples within mathematics instruction [3]. These are consistent with the idea that the effect varies with capacity, but they are summaries rather than full reports of effect sizes.
A replication study focused on working memory capacity and shifting in learning with worked examples, which directly tests whether the capacity-related pattern holds up when the study is repeated [2]. Replication matters here because a single moderation finding can be fragile. The replication study's abstract emphasizes the limited nature of working memory as the reason worked examples matter, which is the same mechanism, but it does not by itself establish that the moderation is large or universal [2].
Where the capacity story is less clear
Not every worked-example study tests working memory capacity. A South African Grade 9 study of faded worked examples looked at gender, socioeconomic status, and prior knowledge, and noted that working memory capacity is predictive of mathematical ability, but it did not analyze capacity as a moderator of the worked-example effect [4]. That means the capacity moderation is supported by some studies but is not a universal feature of every worked-example classroom trial.
The evidence is also thinner for early math learners. A study of worked examples in preschool division tasks examined highlighting correct or incorrect procedures and discussed the limited capacity of working memory, but it did not test whether capacity moderated the effect [5]. So the claim that lower-capacity learners benefit more is best supported for older learners in the studies that directly examined capacity [1][3], and it should not be extended to early childhood without direct evidence.
Sources used in this answer
Does working memory moderate the effect of fading on math performance?
The study investigated faded worked examples in math and examined working memory capacity as a moderator of the effect, finding that capacity plays a role in how much learners benefit.
How working memory capacity and shifting matter for learning with worked examples—A replication study.
A replication study tested working memory capacity and shifting in learning with worked examples, emphasizing that limited working memory is central to why worked examples matter.
Worked-Examples, Prior Knowledge, and Cognitive Load in the Context of Research-Related Instruction
In a review of worked examples, prior knowledge, and cognitive load in mathematics instruction, learners with lower working memory capacity were found to derive more benefit from worked examples.
Incorporating faded worked-examples into the mathematics classroom: exploring gender and socioeconomic effects with two Grade 9 groups from South Africa
A Grade 9 study of faded worked examples analyzed gender, socioeconomic status, and prior knowledge but did not examine working memory capacity as a moderator.
The Role of Worked Examples in Supporting Early Math Learning
A preschool study of worked examples in division tasks examined highlighting correct or incorrect procedures and discussed working memory limits but did not test capacity as a moderator.