Cognitive Load, Memory, and Instruction
by Rebecca Johnson, instructional design researcher and consultant, Innovative Learning Institute
If you’ve walked across a parking lot toward the kiosk, repeating your license plate number under your breath, only to have to turn back to your car and refresh your memory because an ambulance drove by, you’ve experienced Cognitive Load Theory (CLT). How does CLT play out in the classroom? To understand that, we'll need to think first about memory and how we learn.
Learning is built from three types of memory: sensory memory, working memory, and long-term memory. We receive information in the form of sensory input. We select what seems important from that sensory input by using our working memory (or short-term memory). We consider the information in working memory in terms of what we know about the subject already. Some of that working memory can be described as "learned" when it is integrated into long-term memory. There’s research to suggest that frequent practice retrieving information from memory strengthens one’s ability to retrieve that information in the future. A surplus of sensory input at the beginning of the learning process can interfere with working memory. In fact, there’s an inverse relationship between working memory and the amount of sensory input you’re processing at any time (Weinschenk, 2011, p. 47). That said, we certainly don’t need to remember everything—the price of this morning’s coffee, the precise sequence of actions that you took to fill up your car with gas. As Brown et al describe in their book, Make It Stick,
Learning, remembering, and forgetting work together in interesting ways. Durable, robust learning requires that we do two things. First, as we recode and consolidate new material from short-term memory into long-term memory, we must anchor it there securely. Second, we must associate the material with a diverse set of cues that will make us adept at recalling the knowledge later.(Brown, Roediger, & McDaniel, 2014, p. 75)
An understanding of the three types of memory provides an important backdrop for what occurs in Cognitive Load Theory, which describes three types of mental load in any learning experience. Intrinsic loadreflects the difficulty of the content being learned. Germane loadreflects the level or type of cognition you’re asking students to perform (e.g., are you asking them to recall vocabulary words or solve an unstructured problem?). Finally, you’ll want to avoid extraneous load, which is cognitive effort that does not contribute to and even interferes with learning. You might find extraneous load where there’s a surplus of sensory information or poorly written assignment instructions. Cognitive Load Theory reminds us that thinking and learning is work, and that we should be aware of our students working conditions. For the first two types of cognitive load, increasing the level of the intrinsic and/or germane loads should be strategic and should be undertaken in ways that reduce as much extraneous load as possible.
Working with students who have perceptual disabilities requires that instructors think carefully about the limits of sensory memory and about extraneous load. Your DHH students who communicate with ASL are watching you lecture, watching the interpreter, and reading your slides or watching what you write on the board. These students must devote extra cognitive resources to taking in all of this visual information from three different sources, without being able to divide some of the sensory load between vision and hearing. In addition to this huge visual sensory load, DHH students, like all students, are trying to manage the intrinsic load of new content and the germane load of the quality of thinking you’re asking of them. What can you do to help DHH students? The following strategies take into consideration what we know about memory, cognitive load, and instruction.
Signaling or Cueing
While you want to be sensitive to overloading students with visual stimuli, consider how a judicious use of formatting or the use of headings and subheadings can help you communicate course content effectively. Signaling means to use formatting, text, or symbols to draw the learner’s attention to important concepts in a text or in an illustration (Clark & Mayer, 2011, pp. 172–3). This strategy increases intrinsic load and helps learners manage extraneous load by pointing out the most relevant or important information in a document or illustration. My use of italics in the second and third paragraphs of this post is signaling because it serves to highlight important terms. Signaling can also increase germane load by emphasizing the organization of the information and how important terms in the content may relate to one another.
Segmenting or Chunking and Semantic Encoding
There's only so much information that any of us can hold in working memory at any one time. Miller's Law, or "Seven, plus or minus two," is based on research conducted in the early 1950s by George A. Miller at Harvard University. He found that "memory span" or the ability to hold information in one's memory and repeat it back accurately, is limited to seven chunks of information. These "chunks" of information were any grouping of numbers or letters that represents a single meaningful concept to the person being tested. For a mathematician, the number 314159265 could serve as a single chunk of information that is easily remembered and repeated. For those not familiar with the first 9 digits of pi, that number is likely just a string of numbers and would be difficult to repeat back after the first five to seven digits. The ability to recognize and recall information differs among individuals, based on their prior knowledge of any particular subject. The ability to attribute meaning to groups of letters or numbers is called semantic encoding--semantic means "meaning." It's how mathematicians recognize long strings of numbers as pi. It's why we create mnemonic devices. My Very Excellent Mother Just Made Us Noodles (or Nine Pies, if you're having a hard time letting go of Pluto) and variations on this mnemonic helps students remember the order of planets in the galaxy. Stories serve a similar function. A simple story that illustrates the significance of a theorem or a scientific discovery can help students attribute meaning to this new concept. The idea of chunks of knowledge and how semantic encoding aids learning refers back to the "anchors" required for successful consolidation of information into long-term memory in the Brown quote above.
By considering how course content as a whole divides into meaningful subsections, and how even more granularly you might "chunk" the information in your lectures, you can use what we know about learning to help students engage with course content more effectively. When course content is "chunked," students are better able to take in information and learn how it may be related to previously learned topics, emphasizing intrinsic load. Chunking also allows the instructor to remind students about key concepts and how they relate to the structure of the immediate content and to the content of the course as a whole. Chunking, like cueing, increases germane cognitive load. Instructors who develop online courses often “chunk” their material into modules by week or by theme. Another application of chunking is the research on optimum length of course videos, which Philip Guo discovered is no more than six minutes (Hazlett, 2013).
When You’re Teaching, Stop to Smell the Roses and Pick Some Weeds
This strategy combines two principles under one labored analogy. First, slow down and smell the roses. Deaf, hard of hearing, and hearing students all benefit when you deliver course content more slowly (from “Top Ten Teaching Tips | NTID Teach2Connect”). Think of slowing down your delivery as hedging against overload. Give students opportunities in your class to stop, reflect, and consider what they’ve learned. Second, pick the weeds in your presentation or video. Reduce sensory and extraneous load by removing those elements of a presentation or a recorded lecture that are just for decoration--reduce the number of fonts to no more than two, use bold and italics only when necessary, be sure to use high-contrast colors (or black and white), and remove needless animations and complex backgrounds. Reduce the number of words in your presentations. If the entire quote is worth including on a slide, go back to the other advice in this paragraph. Slow down! Give students time to read the entire quote before you resume lecturing.
Remember when I mentioned that retrieving information can strengthen your memory of that information? Research shows that one of the most frequently used study strategies, re-reading and highlighting a text, is ineffective. "Re-reading has three strikes against it. It is time consuming. It doesn't result in durable memory. And it often involves a kind of unwitting self-deception, as growing familiarity with the text comes to feel like mastery of the content." (Brown et al., 2014, p. 10) Disabusing students of the idea that highlighting and re-reading are effective is a worthy enterprise all on its own. But what strategy can students adopt to replace re-reading? It turns out that one of the most effective ways to prepare for a test is to take a test, or mutliple tests.
In one experiment, college students studied prose passages on various scientific topics like those taught in college, and then either took an immediate recall test after the initial exposure or restudied the material. After a delay of two days, the students who took the initial test recalled more of the material than those who simply restudied it (68 v. 54 percent), and this advantage was sustained a week later (56 v. 42 percent). Another experiment found that after one week a study-only group showed the most forgetting of what they initially had been able to recall, forgetting 52 percent, compared to a repeated testing group, who forgot only 10 percent (Brown et al., 2014, p. 39).
To take full advantage of the testing effect, testing should be frequent, but it should also be cumulative, meaning that even as students are taking quizzes in week 10, some of the material in the quiz should be from earlier in the course. Ensuring that you're including information from across the course in each test or quiz reflects research that has found that the testing effect works best when the act of information retrieval is difficult, often because some time has elapsed (Brown et al., 2014, pp. 100–1)
Students aren't the only people who suffer cognitive overload. Before you decide to implement ALL of these or any other principles in your course, take this post's advice and slow down. Read what research tells us about changing instructional practices in STEM. Punch line? It's best to think about your own disciplinary practice when implementing instructional change.
Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make It Stick: The Science of Successful Learning(1 edition). Cambridge, Massachusetts: Belknap Press: An Imprint of Harvard University Press.
Clark, R. C., & Mayer, R. E. (2011). e-Learning and the Science of Instruction: Proven Guidelines for Consumers and Designers of Multimedia Learning(3 edition). San Francisco, CA: Pfeiffer.
Hazlett, C. (2013, November 13). Optimal video length for student engagement. Retrieved from http://blog.edx.org/optimal-video-length-student-engagement/?track=blog
Weinschenk, S. (2011). 100 Things Every Designer Needs to Know About People(1 edition). Berkeley, CA: New Riders.