What is cognitive load?
Cognitive load describes the total amount of mental energy that is required to process new information with the rate-limiting step being working memory. Paas and Sweller (2014) compare knowledge that must be acquired to genetic knowledge obtained through evolution. They divide knowledge into primary and secondary information.
Biologically primary knowledge is the knowledge that we have been genetically programmed to acquire. It is usually easy to obtain, requires little to no effort, and can be assimilated subconsciously (Paas & Sweller, 2014). Biologically secondary knowledge usually involves the use of primary knowledge, it is more difficult to obtain, and energy must be expended to achieve it. As an example, listening and learning your native language is primary, and reading to learn is secondary knowledge.
According to Paas and Sweller (2014), biologically primary information is stored in independent modules in long-term memory, while secondary knowledge is organized into schemata. The more variations of a given related schema a person can process and store, the more fluent they will become in a given subject area or skill.
The architecture of cognitive load
According to Pass and Sweller (2014), biologically primary and secondary information enter working memory to be processed. Biologically primary information is quickly and unconsciously processed. Biologically secondary knowledge is the knowledge referred to in the principles below. Working memory integrates the various components of information into memory models or knowledge constructions. These new memory models are then organized and combined with knowledge in long-term memory to form schema. There are restrictions on the capacity for new information in working memory, but not for information drawn from long-term memory.
Information Store Principle: Long-term memory provides the ability to store an almost unlimited amount of information in the human brain. This information is gathered over a lifetime. Learning something new results in an alteration in long-term memory. Learning results in altering the biologically secondary data that is already stored in long-term memory. As the store of biologically secondary information increases, the user gains fluency in the topic (Paas & Sweller, 2014).
Borrowing and Reorganizing Principle: If each person had to acquire information on their own, it would be difficult to impossible to make any progress in society. To facilitate increasing our secondary knowledge stores, we get it from other people, either by listening to them, reading what they write or by imitating them. Paas and Sweller’s (2014) Cognitive Load Theory assumes that the purpose of instruction then is to assist people in gaining information from other people and, just as importantly, aid them in organizing this knowledge in long-term memory.
Randomness as Genesis Principle: If a learner is not guided when learning something new, then their only option is to make random choices and test them in much the same way that random genetic mutations occur and are tested by the environment (Paas & Sweller, 2014). Each time a learner makes a choice, as long as the options are limited, they could learn from their decisions and be more directed in their learning. Evolutionary unlimited choices could lead to the death of an organism due to indecision.
Narrow Limits of Change Principle: If there are an unlimited number of possibilities presented to a learner, it would be impossible to choose, test, and learn from the choice. Pattern recognition would be nearly impossible. Working memory has both a size limitation, about seven chunks of information, as well as a memory span limitation, 20 seconds without rehearsal, so it serves as a rate-limiting step. The added benefit of this rate-limiting step is that it makes it difficult for the learner to make substantial changes in long-term memory, possibly making it unusable (Paas & Sweller, 2014).
The Environmental Organizing and Linking Principle: Information stored in long-term memory would have little value if it were not correlated with environmental conditions to generate actions. New information entering working memory from the senses is subject to the limitations described above, but information coming from long-term memory is not (Paas & Sweller, 2014).
Three types of cognitive load
Paas and Sweller (2014) define three sources of cognitive load in their theory: intrinsic cognitive load, extraneous cognitive load, and germane cognitive load.
Intrinsic cognitive load: Intrinsic cognitive load is the demand on working memory required to work with the secondary knowledge. It is in direct relation to the complexity of the material itself and is due to the level of interactivity between the knowledge components. The necessary level of energy is determined by the content and cannot be changed (Paas & Sweller, 2014).
Extraneous cognitive load: The effort that is required for working memory to process the information based on the way it is presented. Poor instructional design will increase the number of interacting elements and therefore increase the cognitive load (Pass & Sweller, 2014).
Germane cognitive load: The energy that is left to process the secondary information or intrinsic load into schemata that will be stored in long-term memory. The goal is to decrease extraneous cognitive load and maximize germane cognitive load (Paas & Sweller, 2014).
Assumptions of the Cognitive Theory of Multimedia Learning
Dual Channels: According to Paas and Sweller (2014), humans can process visually/spatially represented material, and auditorily/verbally expressed content in separate channels. If there is sufficient energy available, information entering one channel may be converted to a representation in the other channel (Mayer, 2014).
Limited Capacity: Each of the two channels has a limit to the amount of information that it can process in a given period. A memory span test is one way to test an individual learner’s capacity. A vital function of the executive control part of the cerebral cortex is to allocate resources or determine processing priorities (Mayer, 2014).
Active Processing: Leaning is a dynamic process that requires the ability to select relevant information, organize the data into mental models, and then integrate it with information in long-term memory to form schemata (Mayer, 2014).
Overall premises of the Cognitive Theory of Multimedia Learning
According to Mayer (2014), a multimedia design must not only have a coherent structure, but it must also guide the learner in how best to build a knowledge structure or mental model. The first step is guiding the learner while they select relevant words and images. Then they must help the learner organize the material by allowing them to see relationships between the components and recognize patterns. Finally, the designer must help them build connections between new knowledge and already learned knowledge stored in long-term memory (Mayer, 2014).
Comparing and Contrasting Cognitive Load Theory and the Cognitive Theory of Multimedia Learning
Both Paas and Sweller’s (2014) Cognitive Load Theory (CLT) and Mayer’s Cognitive Theory of Multimedia Learning (CTML) (2014) recognize that there are three types of processing demands or loads on working memory: intrinsic or essential, extraneous, and germane or generative.
In CLT, knowledge is defined as either primary biological knowledge or secondary biological knowledge. The transmission between generations, ease of acquiring the information, and the conscious effort required are used to define them (Pass & Sweller, 2014). Whereas, in CTML, the focus is on the channel that the information enters, whether visual or auditory.
CLT describes the processing of information in terms of cognitive architecture consisting of working memory and long-term memory and applies the following principles: linked to environmental demands, randomly generated, acquired from others, and limited by working memory processing and energy limits (Pass & Sweller, 2014). While Mayer (2014), in CTML, starts with sensory memory, which perceives information and holds it for a brief period to allow the selection of words and images, transduces it, and makes it available to working memory. The information then moves to working memory where seven plus/minus two chunks of information are organized into mental models and integrated with previously learned information (Mayer, 2014).
Long-term memory in both CLT and CTML functions to store information in a highly structured manner. Schema is continually being reorganized and rebuilt into schemata, with the goal being automatic recall.
Similarities Between the Two Models
Both CLT and CTML discuss the structure of human cognitive architecture with CTML adding the sensory phase to working memory and long-term memory (Mayer, 2014). Effective instructional design is used to minimize cognitive load to maximize the limited capacity of working memory. According to Mayer, the end goal is to decrease the amount of necessary extraneous processing so that generative processing can be increased or encouraged (Mayer, 2014).
Methods to Reduce Extraneous Cognitive Load
Redundancy Principle: The redundancy principle states that people learn more deeply when they have graphics and narration rather than graphics, narrative, and on-screen text. Requiring the learner to try to coordinate redundant information with essential information increases memory load and interferes with learning (Pass & Sweller, 2014). The cognitive theory of multimedia learning posits that there are two channels, each with limited processing power, to bring information into working memory. On-screen text and images or diagrams would both use the visual/spatial channel, whereas spoken narration would use the auditory/ verbal channel. Using spoken text and on-screen pictures or diagrams would make use of both channels without overloading either one.
The redundancy effect was not demonstrated when short segments of text were used (Kalyuga & Sweller, 2014). Redundancy may be beneficial for non-native speakers, learners with hearing disabilities, when there are no graphics, or when technical terms are used (Kalyuga & Sweller, 2014). An important caveat of the redundancy principle is that it does not apply when both spoken and written text are both needed for understanding.
Another example of redundancy presented by Kalyuga and Sweller (2014) is the redundancy of actual equipment. Having an instruction manual to read along with a computer may be redundant. Learners may learn more deeply with just the instruction manual. A possible explanation is that when an instructional manual is used along with the equipment, there may be a feeling that any needed instructions could be looked up on a just-in-time basis in the instruction manual instead of learning the content (Kalyuga & Sweller, 2014).
Coherence Principles: The coherence principle states that people learn more deeply from multimedia when extraneous material is excluded. It is tempting to include stories, videos, and random facts that would be interesting to the learner, but not necessary to understand the content. This extraneous information increases extraneous cognitive load, decreases available energy for essential processing, and can be confusing to the learner (Mayer & Fiorella, 2014).
The detrimental effect of adding extraneous information is even more significant in learners with low working memory capacity, the cognitive load imposed by the task, the interest level of the task, or when the content is presented using a systems-based method instead of a learner-based method.
Signaling Principle: The signaling principle states that people learn more deeply when cues are added that highlight the organization of the essential material or to help them select the most relevant information (van Gog, 2014). These cues can be done textually, by adding underlining or bolding, by using imagery such as arrows or highlights, or auditorily by emphasizing certain words.
In the theory of multimedia learning, there are three steps:
- selecting images and text
- organizing images and text
- integrating the information to build connections or schema.
Signaling helps with selecting images and text. This selection process is necessary for the information to be available to working memory.
Research reviewed by Mayer and Fiorella (2014) showed that novice learners tend to attend first to the most salient features of the content. In addition, they showed that learners performed better on transfer tests after reading a summary of the required content instead of a full-length version.
Effective signaling makes more processing power available to facilitate the germane cognitive load. Mayer and Fiorella (2014) also demonstrated that picture cuing had a significant effect on retention, but not on information transfer. Signaling seems to have a more substantial impact on low skill learners than on high skill learners.
Some research reviewed by Mayer and Fiorella (2014) showed that some learners did not perform better with signaling, even though eye-tracking showed that they responded to it. In contrast, learners with more knowledge of the task look faster and proportionately longer at relevant aspects of a task (van Gog, 2014). Signaling is most effective when used with learners who have less background knowledge, when it is used sparingly, and when the display is complex.
Using Refutation Text to Facilitate Conceptual Change
Conceptual change models emphasize the need to determine what the learner’s preconceived ideas are, show that how they are inconsistent or incorrect, and help them to develop a new mental model that is more consistent with scientific evidence. The first step for the learner is to uncover these inconsistencies (Sinatra, Kienhues & Hofer, 2014). If a learner has very little preexisting knowledge, it is relatively easy for them to develop mental models that assimilate the new information into their previous understandings (Posner, Strike, Hewson & Gertzog, 1982).
On the other hand, if the learner has plenty of preexisting beliefs and perceptions, it may be much more difficult to get them to challenge these beliefs. Overcoming this challenge is called accommodation (Posner et al., 1982).
Posner et al. (1982) describe four conditions that must be present for successful accommodation of new knowledge:
- the learner must be dissatisfied with their current understandings;
- they must be able to understand the new information;
- the new information that they have learned must make sense to the learner;
- the new explanation must be fruitful or explain concepts or inconsistencies that their current knowledge structure or schema is not able to (Pintrich et al., 1993).
According to Tippett (2010), learners have three options when faced with scientific knowledge which runs counter to their current beliefs:
- they can reject the new knowledge
- they can memorize the new knowledge and have fragmented knowledge structures
- they can restructure their current schema to incorporate the new information (Pintrich et al., 1993).
One way that conceptual change could be facilitated in the classroom is the use of refutation text. According to Tippett (2010), refutation text is intentionally structured to challenge the learner’s current beliefs. Refutation text starts with a statement of the misconception. It then goes on to state that this conception or belief is not valid. Finally, there is a refutation of the misunderstanding with a statement of the currently understood scientific explanation (Tippett, 2010).
References
Kalyuga, S., & Sweller, J. (2014). The redundancy principle in multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed.) (p. 247-261). New York, NY: Cambridge University Press.
Mayer, R. E. (2014). Cognitive theory of multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed.) (p. 43-71). New York, NY: Cambridge University Press.
Mayer, R. E., & Fiorella, L. (2014). Principles for reducing extraneous processing in multimedia learning: Coherence, signaling, redundancy, spatial contiguity, and temporal contiguity principles. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed.) (p. 279-309). New York, NY: Cambridge University Press.
Paas, F., & Sweller, J. (2014). Implications of cognitive load theory for multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed.) (p. 27-42). New York, NY: Cambridge University Press.
Pintrich, P., Marx, R., & Boyle, R. (1993). Beyond cold conceptual change: The role of motivational beliefs and classroom contextual factors in the process of conceptual change. Review of Educational Research, 63(2), 167-199.
Posner, G., Strike, K., Hewson, P., & Gertzog, W. (1982). Accommodation of a scientific conception: Toward a theory of conceptual change. Science Education, 66, 211-227.
Sinatra, G. M., Kienhues, D., & Hofer, B. K. (2014). Addressing challenges to public understanding of science: Epistemic cognition, motivated reasoning, and conceptual change. Educational Psychologist, 49(2), 123-138.
Tippett, C. D. (2010). Refutation text in science education: A review of two decades of research. International Journal of Science and Mathematics Education, 8, 951-970
van Cog, T. (2014). The signaling (or cueing) principle in multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed.) (p. 263-278). New York, NY: Cambridge University Press.
