Cognitive Load Theory Reconsidered

For our purposes, cognitive load refers to the processing demands placed on working memory capacity during learning. According to the triarchic theory of cognitive load (Sweller, 1999, 2005), there are three types of processing that contribute to cognitive load:

  • intrinsic processing – cognitive processing that is essential for the understanding the learning content and is a result of the complexity of the material and the prior knowledge and expertise of the learner;
  • extraneous processing – processing that does not directly support learning but is a result of the inherent complexity of the learning environment and interface;
  • germane processing – processing that allows the learner to organize the learning content and integrate it with prior knowledge and experience.

Total cognitive load cannot exceed the limits of working memory capacity. Learning is promoted when instructional design minimizes extraneous load in order to free up resources for intrinsic and germane processing, both of which play an essential role in learning. Cognitive load theory (CLT) is robust in that it is supported by reams of evidence, and successfully predicts outcomes particularly for novice learners.

For more experienced and expert learners, CLT is less effective in explaining and predicting learning outcomes. One explanation for this is the “expertise reversal effect”, where designs that promote more efficient learning for beginners, actually inhibit learning for learners with more prior knowledge or experience in a given field (Kalyuga et al. 2003). Kalyuga et al. (2003) suggest that instructional designs that provide support for novice learners, may actually be redundant for experts and as such create additional extraneous processing.

A related explanation for the expertise reversal effect incorporates automatic processing into the the cognitive load equation. Automatic processing is mental processing that occurs without conscious effort or prompting and as such places no demand on working memory resources. Automatic processing is usually the result of extended exposure to material, repetition, and deliberate practice. The more information that can be processed automatically, the less that will contribute to cognitive load. An expert will have routines automated and hence will complete tasks without placing demands on working memory. As soon as novics supports are used, the expert is forced to work intentionally – like a novice, now placing demands on working memory resources.

Challenges to Cognitive Load Theory:

  • Measurement issues – it not clear that existing methods of measuring cognitive load theory are sufficiently discriminating, that is they don’t seem to distinguish between the three different types of cognitive load sufficiently. (see de Leeuw & Mayer 2008)
  • Are the three described types of CL actually distinct from one another (the measurement issues described above suggest ICL & ECL conflated)? (see de Leeuw & Mayer 2008)
  • As noted abouve, the expertise reversal effect is a documented challenge to CLT, and attempts have been made to account for this phenomenon. (see Kalyuga, Ayres, Chandler, & Sweller 2003)
  • Learning of complex schema requires complex learning, which  implies that germane processing follows from, rather than is independent of intrinsic processing. (see van Merriënboer & Sweller 2005)
  • Most recently, “Productive Failure” research calls into question the assumption that because the learner failed the learning task no learning took place. In fact, Kapur (2008) demonstrates that in “productive failure” settings task failure can actually lead to deep meaningful learning.

The new model I propose envisages learner working memory in two parts (instead of three):

  • Attentional – holds data from senses and long term memory.
  • Executive – performs cognitive processes: manipulation, organization, sense making of items in Attentional memory. In other words schema building.


Demands on Attentional resources determined by two variables:

  • Instructional materials (load potential) have the potentialof placing a demand on working memory. This potential demand can result from one of two sources or both:
    • Form – the form of the instructional material is what has traditionally been called the instructional design and refers to content placement, screen or page elements, number of elements on a screen or page, irrelevant material included in the screen or page;
    • Content – the actual content can range from simple to complex and has the potential to tax the working memory accordingly.
  • Learner characteristics (load capacity) refer to the capacity of learners to manage various loads resulting from form or content. Cognitive overloadoccurs when load potential from learning materials exceeds load capacity of the learner.
    • Novice learners will not have great capacity for complex learning forms complex or learning content or both.
    • Experienced will have greater capacity for complex learning forms and content usually resulting from familiarity with the learning form or content.

Instructional material form and content should be tailored to Learner characteristics (ie potential = capacity)

Leave a Reply

Your email address will not be published. Required fields are marked *