A new paper has been published by Ouhao Chen and colleagues that points to a flaw in one of the assumptions of Cognitive Load Theory (CLT); a flaw with some potentially interesting implications.
One of the central ideas of CLT is that the working memory in which we consciously process new (biologically secondary) information is limited, but these limits fall away once dealing with knowledge stored in long term memory. I used to think of this as a design flaw; that you could perhaps improve on humans by expanding our working memories but CLT assumes there is a reason for this limit: Given that information passes through working memory into long term memory, it prevents rapid and disruptive changes to long term memory.
Apart from when dealing with knowledge stored in long term memory, CLT assumes that the working memory capacity of an individual is fixed. This may be wrong because the Chen et. al. paper suggests a new ‘depletion effect’ where working memory capacity reduces further after a period of use.
The authors point to a few previous studies that show that a taxing task led to an impaired performance on a subsequent task designed to measure working memory capacity. However, this was not in a learning context. So the authors set up a context where primary school students learnt about fractions or basic algebra.
All students completed a series of learning tasks, a series of working memory tasks and a test. Roughly half completed this in one massed block whereas the others completed this is spaced-out blocks. Interestingly, the test for this latter group was on a different day to any of the learning tasks and we might think would impair performance even if it led to better learning in the long term.
Not so. Instead, there was a classic spaced practice effect where students in the spaced group did better on the test than those who completed the single massed block. So far, this is unremarkable although the authors do note that relatively few experiments have tested the effect of spaced practice in an authentic classroom environment with authentic learning goals in the way that they did.
But that’s not the interesting part. They also noticed a poorer performance on the working memory tasks for the massed group. Previously, explanations of the effect of spaced practice have relied on ideas such as that the process of trying to retrieve something that is starting to be forgotten improves its retrieval strength. This study suggests another, possibly additional, mechanism: spaced practice avoids cognitive depletion.
The study was quasi-experimental. Rather than being randomly selected into the conditions, for practical reasons children completed them in their usual classes. This was mitigated slightly in the second experiment by running two sequences and switching the conditions from the first to the second sequence i.e the class that did massed practice in the first sequence did spaced practice in the second and vice versa. The results held.
This potentially expands the scope of CLT to act as a theoretical basis to explain other effects like spaced practice.
The authors also note hints in previous experiments that depletion might be highly domain specific. If so, this would have implications for teachers because we could avoid depletion by switching task types.