Submind YouTube summaries
Thumbnail for Week 9 - Lecture 56 : Realtime Feedback From Students

Week 9 - Lecture 56 : Realtime Feedback From Students

Watch on YouTube

Video summary

The lecture focuses on the potential of collecting real-time feedback from students to enhance classroom dynamics and teacher intervention strategies. By treating the classroom as a laboratory environment where students interact with computer-based systems, educators can utilize dashboards to monitor individual struggles and cognitive engagement instantly. When such data is available, teachers can immediately identify students who are "wheel spinning" without progress or those stuck on specific problems, allowing for timely interventions. This approach shifts pedagogy from a static model to an adaptive one where the instructor can provide personalized support, introduce new content, or guide struggling learners based on live data rather than waiting until after the class concludes. To implement this vision, the speaker discusses various existing tools that facilitate immediate data collection, though notes that most are paid services not intended for promotion. Mentimeter is highlighted as a versatile platform for gathering responses to multiple-choice, subjective, or objective questions in real time, while Kahoot offers a gamified experience suitable for smaller classrooms. Additionally, H5P-based methods and Padlet are mentioned as effective ways to pause videos for interactive questioning or collect anonymous feedback, respectively. However, the speaker argues that these standard tools often fall short when it comes to capturing deeper student moods, emotions, or complex cognitive states, suggesting that a more integrated system is needed to truly understand the holistic classroom experience. Given the limitations of current commercial software and the lack of a unified dashboard that combines engagement metrics with affective states, the lecture proposes developing custom solutions using recent advancements in Artificial Intelligence and Large Language Models (LLMs). Teachers or students can leverage these technologies to create bespoke tools that generate QR codes for easy access, utilize cameras and microphones to analyze facial expressions and interactions, and consolidate all data into a single interface. A practical strategy suggested is to engage advanced students as interns or project partners to build these innovative systems, turning their desire to create meaningful software into a solution for educational challenges. Ultimately, the video concludes by emphasizing that while mobile phones have replaced traditional clickers as the primary device for student interaction, there is still no comprehensive system available that fully integrates real-time engagement data with emotional intelligence for teacher use. The speaker encourages educators and students to view this gap as an opportunity to build their own projects using accessible AI frameworks like Coder or cloud-based platforms. By fostering a culture where students help develop these tools, the educational community can create powerful, personalized systems that transform how feedback is collected and acted upon, ensuring that every student's needs are met dynamically during the learning process.
Read the full video transcript
[bell] [music] [bell and music] >> Welcome back. In this video we'll talk about real-time feedback from students. The one thing we saw in the last video is that how to collect, you know, students cognitive engagement of ex-students from using computer vision. Can we collect data from students using, you know, real-time feedback? So, the tablet or the or the desktop kind of system can be kept in the student teacher dashboard, teachers classroom. If you get all the real-time data from the students, students struggle, the problem and everything, a teacher can get a nice dashboard. Then teacher can give feedback. In a laboratory experiment, consider the classroom to lab where the students are interacting with a computer-based system where they are working on a scientific problem. If all the students interaction data is used to provide some feedback to teacher saying that hey, this particular student having struggled in this particular topic. Can you go and intervene the student A is wheel spinning so like trying to go around but not able to achieve the target. Other student B needs help because he's struggling, he's stuck on a problem, he's not able to move on. Now the pedagogy is from teacher. Teacher thinks, "Oh, I know what to do. I could give a feedback. I could give a, you know, new content. I can go and intervene and say what happens. How do I use support you?" Something like that is possible. So, we are talking about can we have a dashboard, a system to the teacher where teacher collects all data and shows it. How do we do that? It's a it's pretty tricky. To do that we'll see. You can think of how do we collect data from the students in the classroom? Can you Can you think of how do we collect data from a feedback from students in the classroom in a large classroom, say 60 students? Write down all those answers. Uh, please pause this video then continue. So, we are not going to go into each of these tools because each one of them is paid and we're not promoting any one of the tool. Uh, it's not our tool to demonstrate. But, I'm going to give you some of the tools of most of you might know Mentimeter. It's the data to collect the, you know, students feedback or students response to particular questions in a real time. You could create a slide with a uh set of questions, you know, um some options questions like MCQ, subjective, objective, you know, there can be a mix and match, polling, voting, all those things. Get the students feedback immediately in your screen. All students need a mobile phone. Almost all the students now have a mobile phone. I'm not talking about undergraduates, not the schools. You could use those devices to collect data immediately. The second one will be Kahoot. I think most of you might know Kahoot from NTNU University in Norway. And they are made um interacting as a more gamified. And it's really fun. And uh Kahoot is also going towards paid. My third no, but Kahoot can be used for small school classrooms. Otherwise, you could have a H5P based methods. There's a lot of H5P as we talked long back. It's uh it's a set of framework, set of rules. But, there are some other uh organizations which create H5P based videos where video can be stopped, you know, paused in between and a question can be asked based on the students response. You know, the video can go In video questions have become popular, you know. And it's everywhere in the MOOCs, in LMS, and everywhere. So, you can use any H5P based system. We recommended personal, but it's again it's a paid thing. It's not that we are promoting any of this. You know, I'm just giving some of the examples which could use. The other very classic um kind of premium uh experience is the Padlet. If you want to collect the student feedback immediately, the response on something anonymously, Padlet. Padlet works excellent because students want to give feedback anonymously all the time. You could get the feedback, then you can provide. So, you could use kind of tools. But, if you know these tools are for one particular reason, but I want to understand students particular uh, different mood or emotion or something like that which is beyond this Mentimeter can do. My suggestion is develop your own tool. How to develop your own tool? It's a very interesting point. So, with with the advent of um, with the recent advancement in the AI, that is LLM, creating a tool become very easier. So, you talk to any of the students. Uh, if you are a student, you could create your own tool, go to LLM. Come up with your idea clearly. I want a tool to be collect data from these many students. I want the data to should give a QR code, shouldn't interact, everything. Clear your ideas very clearly, then develop your own tool. That's the first thing. It's very simple and it's doable. The second step would be instead of developing this, um, my second step can be, you know, if you are a teacher, ask a third-year students or you know, your any students um, in your institute to develop students. Students are really interested to create something meaningful to show it to the world, but they are not able to, you know, understand what to create. They always go back to the same four or five existing methods in the, you know, Kaggle or the GitHub. They use same thing. Instead, give them some opportunity to create a new tool. All the students are interested to that. So, suggestion is create your own tool using LLM, like like Coder or anything, you know, the cloud projects, but you could use, um, any of students as an intern to work it. We also develop a couple of tools by, uh, using, you know, students course project and something like that. And we created many tools using interns. Fantastic tools. And in the recent days with um, advancement in elements very easy. So the idea is that we need to have a system. It can be Mentimeter, Kahoot, or something. Students should have a devices to give feedback. Initially, it was a clicker. The idea concept came that we should have a clicker, but now everybody have a mobile phone. Mobile phone replaces everything. Use the mobile phone effectively. So ask every students to answer questions whenever there's a you know feedback from the students or the students responses. Use the camera. Use the microphone to understand students you know interactions and the facial expressions and cognitive engagement. Collect all this data in a single device, then teacher can act as a personalized system. Why we are not showing any system to demonstrate this particular thing? Because we didn't have any system till now which is created for this. So it's very interesting point. We have a lot of tools which would which would collect students feedback like Mentimeter to teachers during the class after the class. But is there a system which integrates the students engagement affect states and the responses together in a nice dashboard and teacher can interact? It's not. So if you're interested, create that as a project. That will be interesting. And you know that will be a good software or something to this community. Okay, thank you. >> [music] [music] [bell] [music]