Learning Snapshots: Cognitive Load Theory

Overview

The Cognitive Load Theory is a powerful tool in a learning designer’s toolbox. It has had many practical implications in my own learning design work. It has help me not have to walk in the dark when making many design decisions. The theory has been developed since 1988 by John Sweller, an Australian educational psychologist. It focuses on how learners’ cognitive resources are used during learning and problem solving and how we can design instructions and learning experiences to best utilise scarce cognitive resources.

Foundations

This theory utilises concepts and principles from Baddeley and Hitch’s working memory model of memory from the 1970s and George Miller’s information processing theory from the 1950s. Essential assumptions include:

• Our working memory has limited capacity.
• Our working memory has two different systems: one for storing visuospatial information (i.e. things we see such as written texts and images) and one for phonological information (i.e. things we hear such as audio narration).
• Our long term memory has unlimited capacity.
• The main goal of learning is the construction and subsequent automation of schemas.

Key Concepts

Below are brief definitions of key concepts and principles that you often encounter when getting to know more about Cognitive Load Theory.

• The cognitive load of a learner at a specific time is the amount of information he/she is attempting to process in working memory. In another way, it is the amount of mental effort exerted in the working memory.
• Working memory (formerly short-term memory) is responsible for processing information.
• Long term memory is where information stored semi-permanently. Long term memory has unlimited capacity.
• There are three types of cognitive load.
• Intrinsic cognitive load is the load that relates to the inherent characteristics of the content to be learned.  It “can be expressed as experienced difficulty of the subject matter” (de Jong, 2010). This difficulty level depends on the following:
  • the number of domain elements and the level of interactivity between elements in the learning content
  • the extent to which learners have to change ontological categories during the learning process
  • the extent to which the content includes concepts that have developing ontological characteristics
  • the amount of conflicting, ambiguous or uncertain information
  • learners’ prior knowledge
• Extraneous cognitive load is the load caused by the ways the content presented to the learner, i.e. the learning design of the content or the instructor’s ineffective teaching methods. This load is not relevant to learning and does not contribute to schema construction and automation. Therefore, reducing extraneous load will help learners.
• Germane cognitive load is the load caused by the learning processes that contribute to schema construction and automation.
• Extraneous load and intrinsic load are dependent on learners’ prior knowledge or experience (e.g. unfamiliar interface leads to higher extraneous load or the intrinsic load for a given task will be higher for learners with lower prior knowledge).
• Schemas: An individual’s knowledge base is organised in mental structures called schemas. Experts have many schemas while novices have no to few schemas.
• It is also assumed that the distinction between intrinsic, extrinsic/extraneous, and germane loads is clear.

Implications for Learning Design

The findings from cognitive load research are very helpful in designing learning experiences for complex or technical challenging topics, where optimising learners’ cognitive load is particularly important.

As we make design decisions, we need to be mindful that our choices will affect how learners’ limited cognitive resources are used. In general, we should design learner-focused learning experiences, as this aligns with how humans process information and learn.

Increase learners’ cognitive capacity

• In many cases, the modality effect principle suggested that where possible we should include both visual and auditory components in the learning materials. This will encourage learners to make use of both information processing channels and thus increase their working memory’s cognitive capacity.
• Free up working memory capacity through chunking:  one way to help learners chunk information is to help them construct schemas and help them automate schemas, help them integrate elements of lower-level schemas with higher-level schemas.

Reduce extraneous load

Eliminating characteristics of learning material that are unnecessary for learning will help students to focus on the learning process that matter.

• Use many worked examples, goal free problems or completion problems instead of traditional/means-end forms of problem solving. In conventional problem solving where learners have no schema-based knowledge, learners usually use a means-end analysis approach. And this approach can interfere with learning and cause substantial extraneous load. Although learners can solve problems, they may not understand the nature of what they need to learn. Research has shown that using a large number of worked examples instead is more effective. An early concern of using only worked examples is that learners may learn passively and not actively build schemas. To counter this, we can use completion problems where we provide learners with partially completed worked examples and then learners need to generate some key solution steps to complete the problems.
• When suitable, use diagrams to illustrate ideas.
• Don’t present distracting information or features.
• In many cases (but not all), avoid or reduce redundant information.
• Don’t make a task more complex than it needs to be.
• Integrate related sources of information (especially sources that require simultaneous processing) to avoid split-attention effect. Example: when related text and diagram are placed separately, learners have to integrate these sources mentally, and this leads to unnecessary heavy cognitive load.
• Provide clear instructions, e.g. clarify text with examples.
• Consider visual aspects of the learning experience: for example, using a tiny font size may cause extraneous load when learners need to try hard to read.
• Provide summaries after major learning objectives: for example, we often see a summary after each book section/chapter.

Note that according to the expertise reversal effect, instructional techniques that work for novice learners may not work for learners with high expertise. For example, detailed guidance that helps a novice learner build schema may become redundant information and cause extraneous load for expert learners. Therefore, it is essential to consider learners’ prior knowledge when making learning design decisions.

Maximize germane load

Learning designers should try to encourage learning processes that cause germane (schema related) load by incorporating activities that support schema building and automation. At the same time, we also need to take care that these activities don’t tax working memory capacity too much. Activities that help/stimulate/guide learners focus on cognitive processes that help construct schemas include interpreting, exemplifying, classifying, inferring, differentiating, and organizing (Mayer, 2002).

Manage the intrinsic load

“For a given task with a specific knowledge level, intrinsic cognitive load is fixed.” (Paas & Sweller, 2014)

Whether intrinsic load can be changed or reduced is controversial. There are two different camps on this matter. According to proponents of the view that intrinsic load can be reduced, some ideas to minimise intrinsic load include:

• Break down a complex task into smaller and simpler steps. Example: show steps in drawing an animal.
• Sequence the material in a simple-to-complex order.
• Sequence the material using the part-whole approach, introducing the independent partial elements first before presenting the whole integrated tasks.
• Sequence the material using the whole-part approach, delivering the entire material from the start but focus learners’ attention only on different subsets of elements.

Controversies and Challenges

Following are some questions that have emerged:

How do we balance between reducing extraneous load and encouraging germane load?

Same design features may increase both extraneous load and germane load. Let’s look at the same example above of placing related text and diagram separately. In this case, learners have to integrate these sources mentally, which increases the cognitive load. Some of that load is unnecessary (extraneous load), and some of that requires learners to think deeply (germane load).   What should we do in those cases?

How do learners’ prior experience affect our design decisions?

These loads are not independent of learners’ prior experience. We need to consider learners’ previous experience when managing different types of load. What works for novice learners might not work for expert learners.

What aspects of learning materials are extraneous?

It is not always evident to know which characteristics or aspects of learning materials are extraneous.

Can extraneous load ever be zero?

There seems to be a limit in the reduction of extraneous cognitive load. It would be challenging to create learning materials with no extraneous load.

Can intrinsic cognitive load be reduced?

There are two different views on this, however, it mainly depends on how intrinsic load is defined. The original thought is that intrinsic cognitive load can’t be reduced.

Can germane load ever be too high?

Yes, and when it too high, it can become a form of extraneous load and inhibit learning. For example, in a study, Stull and Mayer (2007) found that students who used existing graphic organisers outperformed students who created these organisers themselves.

Conclusion

Managing the different cognitive loads effectively will enable learners’ cognitive resources to be used effectively in their learning processes.

References

Baddeley, A. D., & Hitch, G. (1974). Working Memory. Psychology of Learning and Motivation, 8, 47-89. https://www.sciencedirect.com/science/article/pii/S0079742108604521

Chandler, P., & Sweller, J. (1991). Cognitive Load Theory and the Format of Instruction. Cognition and Instruction, 8, 293-332. https://doi.org/10.1207/s1532690xci0804_2

Clark, R. C., Nguyen, F., & Sweller, J. (2005). Efficiency in learning: Evidence-based guidelines to manage cognitive load. Pfeiffer.

de Jong, T. (2010). Cognitive load theory, educational research, and instructional design: Some food for thought. Instructional Science, 38, 105-134. https://doi.org/10.1007/s11251-009-9110-0

Information Processing Theory – IResearchNet. http://psychology.iresearchnet.com/developmental-psychology/cognitive-development/information-processing-theory/

Kalyuga, S. (2007). Expertise reversal effect and its implications for learner-tailored instruction. Educational Psychology Review, 19, 509-539. https://doi.org/10.1007/s10648-007-9054-3

Mayer, R. E. (2002). Rote versus meaningful learning. Theory into practice, 41(4), 226-232. https://www.jstor.org/stable/pdf/1477407.pdf

Paas, F., & Sweller, J. (2014). Implications of cognitive load theory for multimedia learning.  In The Cambridge Handbook of Multimedia Learning, Second Edition. https://doi.org/10.1017/CBO9781139547369.004

Stull, A. T., & Mayer, R. E. (2007). Learning by Doing Versus Learning by Viewing: Three Experimental Comparisons of Learner-Generated Versus Author-Provided Graphic Organizers. Journal of Educational Psychology, 99, 808-820. https://doi.org/10.1037/0022-0663.99.4.808

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive science, 12(2), 257-285. https://onlinelibrary.wiley.com/doi/pdf/10.1207/s15516709cog1202_4