In the evolving landscape of educational technology, the gap between AI capabilities and collaborative learning has been a persistent challenge. Today, YoChatGPT! bridges this gap with our innovative polling feature, enabling true social constructivism in AI-enhanced learning environments.
Our platform uniquely offers multi-user synchronous collaboration, multiple AI model access, and now, interactive polling - all designed to transform how students and teachers engage with AI in education. Recent feedback from our seminars and workshops revealed overwhelming enthusiasm for these collaborative features, with educators particularly highlighting the platform's ability to “facilitate team-based learning and social interaction".
YoChatGPT! is capable of implementing gamified learning activities through context-aware rooms which facilitate interactive formative assessment, such as multiple-choice questions with timed responses and automated point allocation based on response speed. This gamified approach can be applied to problem-based learning activities which enhances student engagement and motivation [1,2].
In addition, we are able to implement Peer Instruction (PI), an evidence-based active teaching pedagogy, within YoChatGPT! to enhance critical analysis of generative AI outputs. PI, as described by Mazur and supported by research at Harvard University, significantly improves student engagement and comprehension [3]. Within YoChatGPT!, instructors can pose the traditional conceptual multiple choice questions, supplemented by generative AI responses prompting individual reflection followed by peer discussions and generative AI critical thinking. The platform's voting feature facilitates real-time feedback and assessment, mirroring the benefits of technology-enhanced PI with generative AI.
In this article, we will delve into and explore active learning teaching strategies incorporated with generative AI. In particular, we will implement two active learning pedagogies: (i) gamified problem-based learning and (ii) peer instruction incorporated with generative AI. More precisely, we will show how to use YoChatGPT! voting functionality to implement peer instruction teaching strategy. For gamified problem-based learning, we will use YoChatGPT!’s context-aware rooms to calculate a formula which rewards faster, accurate responses, encouraging active participation.
Modified Peer Instruction with Generative Ai to Promote Critical Thinking
At the current moment, existing generative AI platforms are designed for individual use only, hindering classroom integration and collaboration. Moreover, teachers often showcase personal generative AI use, leading to passive observation rather than active student engagement. These passive methods of teaching with genAI led us to develop the voting functionality in YoChatGPT!, to be able to implement an active teaching pedagogy called, Peer Instruction (PI). However, we modify and add to the traditional PI strategy [4] by adding a last additional step, where the instructor shows the generative AI response to the same multiple choice question; and then does a 3rd last vote to see if this changes the distribution of student answers. In this modified generative AI peer instruction strategy, students have each other to discuss whether their choice is correct along with input from the the generative AI. This encourages increased critical thinking in the last step of this modified peer instruction strategy, integrated with generative AI.
Figure 1: Shows the modified PI with generative AI, where in the last step, we provide students with a generative AI response to the question, and ask them to vote a 3rd time to see if the generative AI answer effects their final answer in the end. For the actual room link, please see: https://www.yochatgpt.io/chat_room/Peer%20Instruction
Gamified Problem-Based-Learning with Generative AI
One of the best technologies to implement gamified problem-based-learning may arguably be Kahoot! However, with context-aware rooms on YoChatGPT!, users can ask students to answer normal multiple choice questions by inputting their letter answer (A,B,C, D or E); and then instructors can prompt engineer the LLM to calculate the Kahoot! points according to [5] without any need of hard coding the data collection into the back end of the online web application. Instructors can first pose a multiple choice problem to students, and then put the following prompt into one of YoChatGPT’s LLMs:
Prompt #1: Suppose we are calculating points, just like Kahoot for the fastest correct response, get's most points, according to the following formula [5]: ⌊ ( 1 - (( {response time} / {question timer} ) / 2 )) {points possible} . Please calculate the amount of points of each student in a table, with columns containing the user name, what their answer was, the time they responded, with the last column being the points they obtained, according to the formula above, with question timer being defined as 30 seconds; and response time = (time student answered the question (in seconds)) - (time question was posed in the chatroom by the instructor), and possible points = 1000; and the correct answer being D.
Then, to calculate the points for subsequent questions, the prompt could be:
Prompt #2: Please calculate and update the amount of cumulative points of the above table, given that the correct answer is C for the second question
Figure 2: Shows prompt engineering in YoChatGPT!s rooms to calculate Kahoot points of multiple choice answers. For the actual room, please see link: https://www.yochatgpt.io/chat_room/Kahoot%20Room2
References:
[1] Fukuzawa, S., Boyd, C., & Cahn, J. (2017). Student motivation in response to problem-based learning. Collected Essays on Learning and Teaching, 10, 175-188. https://doi.org/10.22329/celt.v10i0.4748
[2] Antonaci, A., Klemke, R., & Specht, M. (2019). The effects of gamification in online learning environments: a systematic literature review. Informatics, 6(3), 32. https://doi.org/10.3390/informatics6030032
[3] Crouch, C. H., & Mazur, E. (2001). Peer instruction: Ten years of experience and results. American journal of physics, 69(9), 970-977.
[4] Peer Instruction. ABL Connect. Derek Bok Center for Teaching and Learning. Harvard University. Retrieved on Oct 27, 2019, from Peer Instruction for Active Learning https://lsa.umich.edu/technology-services/news-events/all-news/teaching-tip-of-the-week/peer-instruction-for-active-learning.html
[5] Kahoot support, “How points work” https://support.kahoot.com/hc/en-us/articles/115002303908-How-points-work#:~:text=Points%20are%20awarded%20based%20on,time%20by%20the%20question%20timer.
Share this post