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]