Learning-by-teaching is widely recognised as a highly effective educational method. Reflecting on my personal journey in higher education, I recall many instances where the necessity to teach a concept to students or colleagues forced me to immerse myself further into the subject, thereby enriching my own understanding and knowledge. I would also often find that insightful questions, discussions, and the unique perspectives others brought sparked fresh revelations and discoveries for me.
A new and noteworthy study by Wang et al. (2023) puts the spotlight back on this topic and delves into how we might best use this learning-by-teaching method in the digital age.
The authors provide a novel perspective on the teaching method – teaching without an audience. The theory behind teaching ‘to no one’ is that the teacher will not feel the pressure, stress and nervousness associated with public speaking and therefore will have fewer distractions from the task of knowledge acquisition.
In this study, students were divided into three groups and were asked to explain a topic by either teaching another student (teach-to-student), a group of students (teach-to-group) or they were asked to make a lecture video (teach-to-camera). The results suggested that teaching to an ‘imaginary audience’ via video is more effective for learning than teaching to a student or a group of students face-to-face. This is mainly due to that method minimising distractions (such as stress, fear, worrying thoughts, anxiety, and nervousness) while freeing up the capacity to increase active generative processing, allowing students to make connections between new information and their existing knowledge.
Interestingly, this form of learning aligns well with the capabilities of Large Language Models such as ChatGPT. So, how can we use ChatGPT to emulate the idea of teaching to an ‘imaginary audience’?
Applying the Learning-by-Teaching Approach with ChatGPT
One scenario we might consider is to ask a student to ‘teach’ a concept they’ve recently studied to ChatGPT (this also works with other LLMs such as Bing or Bard). For example, if a student has just learnt about oxygen consumption and its relation to physical performance, they could be given the task of explaining this topic to ChatGPT.
The student would need guidance on how to start and end the statement they submit to ChatGPT. An example follows:
“ChatGPT, imagine you’re Sam, a third-year Sport and Exercise Science student, and I’m your classmate. Today, we are studying V̇O2 max”
[Student to enter their own explanation of the topic next]
“V̇O2 max, refers to the maximum amount of oxygen a person can utilise during intense exercise. It is measured in millilitres per kilogram of body weight per minute (ml/kg/min) and is an indicator of aerobic fitness. The higher the V̇O2 max, the more oxygen a person can use, and the better their aerobic fitness. Athletes often aim to increase their V̇O2 max because it can enhance endurance and overall performance.”
[Student asked to close with the following question]
“Does it make sense so far, Sam? Would you like to ask me any questions?”
Upon finishing their explanation, ChatGPT will respond with some follow-up questions to simulate a real student’s curiosity or need for clarification. Here is a response ChatGPT provided to our example:
“That’s interesting! Could you delve deeper into how athletes train to increase their V̇O2 max? And what are some factors that might affect a person’s V̇O2 max?”
As part of the task, the student would then need to respond with correct and appropriate information. If their response was still insufficient, ChatGPT might respond with the following:
“Thanks for the explanation, but I’m still not quite clear on a few things. How exactly does the training process help increase V̇O2 max? Are there specific exercises that are more effective? And how do factors like age, gender, and lifestyle influence a person’s V̇O2 max?”
If the student provided an accurate, thorough response, ChatGPT could then acknowledge the understanding, providing positive reinforcement. For example:
“Wow, that’s a great explanation! You’ve clearly explained how specific training methods can help improve V̇O2 max and how various factors can influence it. Thanks for taking the time to teach me, I’ve learned a lot!”
Students could then be asked to submit a copy of their ‘conversation’ with ChatGPT and highlight the parts that provided them with additional knowledge.
This type of questioning will require students to do further research into the topic where their knowledge is currently insufficient, thereby using the teaching process to solidify their understanding. This approach also helps develop their ability to respond to questioning and foster an engaging learning environment, even when they are the ones providing instruction.
The learning-by-teaching approach using AI tools like ChatGPT can create an environment for students to clarify their understanding, identify knowledge gaps, and convey complex information in a simplified manner. Importantly, it eliminates the social pressure often associated with teaching in front of peers, thereby boosting the efficacy of the learning process.
As technology evolves, so must our approaches to education. By combining traditional teaching methods with new tools like ChatGPT, we can enhance learning experiences and better prepare our students for a future steeped in technology.
Teaching to nobody (teach-to-camera) is what makes social media content developers so effective. It is a great way to learn how to deliver content effectively, as you are likely to be the most thorough critic of yourself.