Uncovering Hidden knowledge

It took me about 20 minutes and a lot of searching like this: “act as an expert in engagement in learning. Help me to find a researcher from Australian or New Zealand who found that students enter a classroom knowing about 50% of the content to be taught,” before I remembered the name of the researcher I wanted to write about today, Graham Nuthall. AI is helpful, but sometimes it’s your own prior knowledge and a little time (and grace) that allows you to pull prior knowledge from the recesses of your brain. Coincidentally, Nuthall’s work is about prior knowledge. His research offers a roadmap to understanding how our students learn and how we can optimize our teaching approaches to meet their needs.

Graham Nuthall’s research revealed a mind-boggling truth: when we start a lesson, half of what we are about to teach is already known by our students. The trick is that each student holds a different part of the puzzle. Because of this, Nuthall discovered that almost every student learns something different in a lesson and they often learn more from each other than from the teacher, which can include both accurate information and misconceptions. This dynamic makes our role as discoverers of truth and facilitators of deep learning for each student critical.

In essence, Nuthall highlighted that students enter lessons with vastly differing levels of prior knowledge and preconceptions about a topic. Accurately assessing and addressing these prior knowledge states is crucial for effective teaching and learning to occur. 

Furthermore:

• Nuthall’s research showed that a student’s level of prior knowledge was a much stronger predictor of their learning than factors like ethnicity or measured intelligence.

• Students interpret classroom activities and new information based on their own existing goals, interests, and background knowledge. Their prior knowledge significantly shapes what they extract and learn from the lessons.

• Students often have prior misconceptions or incomplete understandings about concepts before they are formally taught. These pre-existing ideas can hinder their ability to grasp the correct concepts being taught.

• Students with little or no relevant prior knowledge tended to learn more from direct teacher instruction and explanations.

• Students who already had substantial knowledge about a topic learn more by working independently or with peers on related activities and projects.

• Activating and connecting new learning to students’ existing relevant prior knowledge facilitated the integration and retention of new concepts through the process of subsumption (essentially adding new knowledge into a current mental model).

The implications of all of the above findings is that teachers need to be aware of the wide range of prior knowledge their students possess and tailor instruction accordingly – leveraging accurate prior knowledge while re-shaping inaccurate preconceptions through carefully designed learning experiences. It goes without saying that one learning experience will not work for all learners.

This is all great in theory, but having the ability to do this practically, like many other educational theories, has been hard to imagine for many teachers with large class sizes and a year’s worth of content to cover (not to mention other life and social skills that teachers are endeavoring to incorporate in their lessons). This challenge is much more accessible now with AI tools. 

Assessing Prior Knowledge: At the beginning of a unit, we can ask students to write down everything they know about the topic we are about to explore. (Students could also use speech to text apps if that’s a more accessible way to share their learning or a translation tool if they are learning in a language that they are not yet fluent in.) Using ChatGPT, Claude, Gemini, or any other AI that’s available, we can prompt the AI to analyze each student’s prior knowledge based on the concepts, content, and skills that we are about to teach. 

Grouping: Ask the AI to group students based on their prior knowledge. Then ask it to suggest groupings for different concepts, content, and skills so that groupings are flexible depending on the learning outcomes.

Differentiation: From there, it’s possible to develop direct instruction mini lessons for students with less prior knowledge and project-based learning for students who already have a foundation in the topic. (We can use AI to help develop these as well).

Formative Assessment: With deft prompting and rubrics, we can use AI to help give feedback to students on their work and assess whether students have understood the concepts being uncovered or built upon in the unit. This frees up time for us to work with students who may need additional support in accessing content, concepts, or skills.

Ultimately, Nuthall’s research serves as a reminder that effective teaching requires more than just delivering content; it involves carefully crafting learning experiences that build upon students’ existing knowledge, addressing their misconceptions, and providing opportunities for meaningful construction of understanding. By embracing these principles and leveraging the power of AI, we can create classrooms that truly meet the diverse needs of our students, fostering a love for learning and equipping them with the skills and knowledge necessary for success.

https://researched.org.uk/2019/02/26/graham-nuthall-educational-research-at-its-best/