I recently explained to a colleague about how engaging with cognitive science has helped me develop as a teacher.
‘Interesting… but what does that look like when I’m teaching Year 9 on a Tuesday afternoon?’
I described a couple of lessons I had taught, and he immediately understood what I meant. He just needed concrete, subject-specific examples. Specialist subject knowledge is important to science teachers (Kind, 2014), and ‘expert teachers’ are able to prioritise this knowledge so that they can respond appropriately to the demands of a lesson (Schempp et al., 2002). In other words, they know how and when to apply key evidence-based strategies on a Tuesday afternoon with 30 pupils in front of them, even if they hadn’t previously planned to use them.
I enjoy teaching science topics that demand a solid conceptual understanding, especially when those underlying concepts are tricky to understand at first, because it is satisfying to see ideas finally clicking into place. I think that one of the most beautiful things you can experience as a teacher is to see a struggling student suddenly ‘get it’. This is why ‘threshold concepts’, which have been described as portals to new or transformed understanding, captured my imagination when I first encountered them (Chandler-Grevatt, 2015). Threshold concepts, by definition, are also:
- troublesome, but when you master them, a range of linked ideas fall into place
- transformative, as opposed to ‘Big Ideas’, which help you answer questions and make predictions but don’t induce profound conceptual change (Talanquer, 2015)
- irreversible, changing your outlook on the world once encountered.
To identify threshold concepts in chemistry, I used misconceptions literature as a starting point to compile a list, then determined the relative strength and pervasiveness of these misconceptions by measuring the ‘confidence difference’ for each one (the difference between the total confidence scores for right and wrong answers) using diagnostic questions with confidence measures. A large confidence difference indicates a relatively high proportion of correct answers with a high degree of certainty from students. I used combined scores from four classes for each concept to ascertain its relative difficulty (Kaiser, 2017).
This meant that I was better able to anticipate misconceptions and conceptual hurdles. For example, dot/cross diagrams can give students the idea that ions have a localised charge (Kind, 2004), whereas charge is actually distributed throughout simple ions, acting in all directions. I found that this misconception was particularly persistent among my students, so I made sure that I addressed it specifically.
However, identifying threshold concepts is not straightforward, because we first need to ‘retrace the journey back to innocence’ (Cousin, 2006), which is tricky when you’re an expert (as teachers are). For example, one idea I think might qualify as a threshold concept is that lab-scale reactions involve millions of particles, not just the few particles that are drawn in dot/cross diagrams. This took me by surprise because it is obvious to me, but it became apparent when I was talking to students that they didn’t necessarily realise this. Our tacit knowledge is not necessarily evident to our students unless we specifically point it out, and some misunderstandings may only be uncovered through conversations with our students.
Threshold concepts are (by nature) troublesome, so understanding them can be a liminal process, and students can feel tempted to seek ‘sanctuary’ in mimicry before they have fully mastered a concept (Cousin, 2006). For example, they might give correct answers to conceptual questions (either inadvertently or via a learned short-cut) without fully understanding why their answers are right. But this transition state is an important part of the learning process (Land, 2017), and I have been trying to find ways of supporting students through it.
Some of the approaches I have used to help students to overcome conceptual hurdles came from cognitive science, and last year I helped form the #CogSciSci peer-support network. To date, 120 members have shared 300 forum posts, discussing a wide range of topics on applying cognitive science to science teaching, which has really helped me by giving me concrete examples. For example, most #CogSciSci teachers now use retrieval practice routinely, often via low-stakes quizzing. I use retrieval practice to help students remember key language, partly because I think it will aid conceptual understanding – it will be easier for students to conceptualise processes involving millions of ions if they don’t have to remember what an ion is first.
Students use both working memory (WM) and long-term memory (LTM) to solve problems, and drawing on Cognitive Load Theory, I know it’s important to avoid overloading WM when introducing problems for students to solve. This is especially important in science, where conceptual understanding, mathematical processing and tacit knowledge also tend to be needed. For example, to determine the formula of an ionic compound such as magnesium sulphate, you first need to work out the charges on the ions within the compound. This is a simple step for me, but it involves tacit knowledge, retrieval of memorised information, and procedures that I have carried out so many times that they’re now automated. #CogSciSci teachers routinely use worked examples and ‘shed loads of practice’ (Boxer, 2017) to support mastery and automaticity, and I have used this approach to avoid overloading students’ WM, leaving them with a greater capacity for higher-order processing. I talk through my reasoning as I demonstrate worked examples, making tacit knowledge explicit, so although it’s obvious to me that Mg will form a positive ion, I make sure I think aloud about why it does.
Using subject-specific approaches from cognitive science has helped me to develop my teaching. But I would not have been able to apply them without concrete examples from my peers in #CogSciSci. Without them, this article would have been much more abstract.