Adam Stubbs, Teacher of Science, Park View School, UK
Research into learning and cognition has led to the use of various evidence-based approaches in the classroom. Techniques such as retrieval practice and spaced learning have been shown to increase retention and boost learning (Karpicke and Roediger III, 2007). These are important approaches in supporting the retention of knowledge. But what if they also reinforce misconceptions and errors?
Implicitly reinforcing misconceptions
Misconceptions that are embedded at the start of a topic may be reiterated and amplified as they are repeatedly retrieved. The use of practices such as spaced retrieval can increase retention, but we need to make sure that we are retrieving the desired knowledge, and not perpetuating the development of misconceptions that can arise from our teaching.
Some misconceptions can be embedded implicitly – concepts that are not explicitly mentioned during instruction, but which are reinforced by the design of the wider instructional sequence. For example, in chemistry we teach how ionic salts can dissolve in water to make liquid solutions. It’s possible to design instruction that explains this but perhaps doesn’t mention that ionic salts that have been melted behave similarly. In this case, the instruction is perfectly valid but incomplete, and risks reinforcing the common misconception that all liquids contain water.
We can design instruction to prevent misconceptions forming by managing the sequence of ideas introduced at each stage of instruction. Explanatory sequences are step-by-step explanations that progressively introduce new knowledge. Instead of giving students large volumes of new content at once, or giving explanations that may be perfectly accurate but impractical for a novice learner, consideration is given to ensure that students fully understand new content.
Managing cognitive load
Effective explanatory sequences should account for the learners’ inherently limited working memories. A task with high cognitive load can easily overwhelm novice students, leading to their missing key information or reverting to errors built on prior knowledge. Cognitive load is challenging to measure but, by incorporating knowledge into a learner’s long-term memory, we can reduce the burden on their working memory during a task (Sweller, 2016). For novice learners, instructional sequences should be designed to take this into account, with a focus on integrating initial knowledge to prevent overload. As working memory is limited, it is important to ensure that students are focusing their attention on the most desirable information (Klepsch and Seufert, 2020).
Effective explanations should therefore manage cognitive load to support students in understanding new information without overwhelming them through excessive content. For example, if I’m using a diagram such as that in Figure 1 to introduce and model the process of electrolysis, then there are several questions that I’d consider in my planning:
- What do I want students to be focusing on at each point in my explanation? Considering our limited working memories, this should be as specific as possible.
- Is there anything novel in the diagram or explanation?
- Do students have sufficient background knowledge to fully understand the diagram?
- Could they be distracted by extraneous information or text?
The diagram is simplified as much as possible to ensure that student attention is directed to the desired components at each stage. Using basic structures and simple line-diagrams removes any unnecessary complications and/or ‘seductive details’ (Harp and Mayer, 1998, p. 414). Labels are only added once students have a conceptual understanding, and components are simplified instead of being realistic. Realistic or artistic details may make the diagram prettier, but they can distract students from the conceptual focus. Similarly, interesting side notes are omitted from my verbal explanation to limit distraction from conceptual demands. There is certainly a place for such details, but introducing them later allows students to focus and retain the desired content more readily (Harp and Mayer, 1998).
Is this necessary right now?
Each stage introduces only the minimum necessary information, allowing the teacher to check more precisely for misconceptions. The process of constructing a diagram step by step allows the teacher to explain each stage, whilst managing cognitive load through the direction of student attention (Paivio, 2013). The diagrams in this example begin unlabelled, with labels added for any parts that students are unfamiliar with. During initial instruction, adding excess labels can direct attention away from the overall process being modelled (Jamet and Le Bohec, 2007). For any additional labels or visual components, a decision should be made as to whether they are necessary at this exact stage.
It is necessary that students know that the rods in Figure 1 are called electrodes, but it is not necessary at this point to label the positive and negative electrodes as the anodes and cathodes. Those labels, whilst useful, are not necessary during initial instruction and can be integrated later on.
Minimal necessary variation – directing attention
Explicit instruction is particularly beneficial when new content is introduced, to ensure that student attention is directed to the key concepts in the task by breaking down content and introducing only the minimal necessary information at each stage. This ensures that students focus initially on remembering a few units of information and applying these using simple procedures. The delivery of short explanations on the most necessary content prevents students’ working memories from becoming overloaded, allowing them to actively process the new information more effectively.
As learners become more knowledgeable, the degree of instructional support should decrease; therefore, it is important to consider the amount of support and task variation to include. Task variation allows learners to consolidate more question types within their schemas. However, this must be balanced with cognitive load demands. Engelmann and Carnine (2016) suggest that sequencing can be designed using the concept of minimally different variation. Through worked examples that differ minimally during instruction, learners can more efficiently process sameness and difference within related questions. As learners process each example, minimal variation prevents learners from attributing differences in approach to differences in the surface structure of problems.
For example, Figure 1 shows one possible representation of electrolysis apparatus. If initial instruction showed worked examples of this diagram alongside more complex examples with unusual apparatus, then learners may interpret differences in understanding as the result of differences in surface variation, as opposed to underlying conceptual differences. Minimal variation in question sequencing prevents this misattribution.
Using minimal necessary variation prevents cognitive overload for novice learners, and increases the opportunities for feedback and instructional assessment. A requirement of this approach is to introduce content in small sections, with repeated sequences composed of minimally similar examples. Increasing question frequency provides more opportunities for teacher feedback and increases feedback precision. Just as minimal variation reduces incorrect attribution for learners, it does so for teachers during feedback too, as feedback is more precise when variation is controlled. This maximises opportunities for student success and positive reinforcement, which directly sustains increases in student self-efficacy and subject beliefs (Schunk, 1991).
Instruction requires the teacher to relay information as the subject expert. However, part of this requirement is precluded by the ‘curse of knowledge’, whereby expertise in a subject can lead to teachers misattributing their own fluency within the topic onto more novice learners. This can be actively avoided during instructional planning, through the concept of faultless communications (Engelmann and Carnine, 2016). These are explanations through which the only possible interpretation is that which was intended.
By considering the most common misconceptions in a topic, we can proactively design explanations to prevent misconceptions from developing. A common misconception in electrolysis is that all liquids contain water. Figure 1 shows a five-step sequence, which begins with a solid salt that is melted. By including this first diagram in the sequence, it ensures that the learner is explicitly aware that this process doesn’t require water. By including this early in the sequence, it helps to prevent such a misconception from developing. To reduce the chance of common misconceptions occurring, there are four questions to consider:
- What are the common misconceptions in this topic and related topics?
- Where do students usually make mistakes?
- Does my current explanation encourage or inhibit this?
- How can these be actively prevented?
Explanatory sequences must find the balance between introducing new concepts and maintaining student fluency, whilst considering how and when misconceptions could develop. Questioning our own instructional sequences in response to research evidence makes it possible to design better explanations that both manage cognitive load and help to prevent misconceptions from arising and being reinforced.
Engelmann S and Carnine D (2016) Theoretical foundations. In: Engelmann S (ed) Theory of Instruction: Principles and Applications. Eugene, Oregon: NIFDI Press, pp. 3–10.
Harp SF and Mayer RE (1998) How seductive details do their damage: A theory of cognitive interest in science learning. Journal of Educational Psychology 90(3): 414–434.
Jamet E and Le Bohec O (2007) The effect of redundant text in multimedia instruction. Contemporary Educational Psychology 32(4): 588–598.
Karpicke JC and Roediger III HL (2007) Expanding retrieval practice promotes short-term retention, but equally spaced retrieval enhances long-term retention. Journal of Experimental Psychology: Learning, Memory, and Cognition 33(4): 704–719.
Klepsch M and Seufert T (2020) Understanding instructional design effects by differentiated measurement of intrinsic, extraneous, and germane cognitive load. Instructional Science 48: 45–77.
Paivio A (2013) Mind and its Evolution: A Dual coding Theoretical Approach. Hove, UK: Psychology Press.
Schunk DH (1991) Self-efficacy and academic motivation. Educational Psychologist 26(3–4): 207–231.
Sweller J (2016) Working memory, long-term memory, and instructional design. Journal of Applied Research in Memory and Cognition 5(4): 360–367.