With critical nationwide staffing shortages and post-pandemic learning loss to contend with, teachers have less time than ever to devote to honing their classroom skills.
One of the most effective ways teachers can develop insights into their instruction is by watching themselves deliver lessons on video. For example, they get visibility into their student interactions and see which questions elicit the best discussion. The Institute of Education Sciences noted in a recent study that almost 90% of teachers said they were more reflective about their teaching as a result of the feedback they received via video coaching.
We created the Sibme platform in 2013 to ensure professional learning opportunities were efficient and based on concrete evidence of practice. Now, we’ve harnessed the power of artificial intelligence (AI) with our new AI, giving educators a quicker way to gather qualitative and quantitative data related to their teaching. Sibme AI aids teachers, administrators and instructional coaches to identify areas for improvement.
The Path to AI Evolution
As we added new features to Sibme, we continued to believe in the power of video and its ability to provide an accurate view of what’s happening in the classroom. As technology evolved, we recognized ways to make it even easier and faster for teachers and instructional leaders to analyze recorded classroom instruction. Sibme is not an AI platform; rather, we harnessed AI to help educators more quickly identify and analyze the data from Sibme videos.
At its core, Sibme allows educators to use any device to record classroom lessons. These videos then upload to the platform. From our platform, educators can easily crop videos and share them with others. Our original features—video recording, sharing, storage and time-stamped video commenting—continue to give teachers, coaches and administrators a full view of a lesson. This also allows coaches to provide teachers with more actionable and objective feedback.
The platform grew to offer virtual coaching to teachers, instructional coaches and teacher cohorts. Using the videos, coaches examine evidence of the educator’s practice and work with them to develop instructional strategies. Virtual coaches still use the platform to share comments on lesson videos, documents and other insights. With the introduction of Sibme AI, we can now harness the power of artificial intelligence to further enhance professional learning and save educators even more time.
The Progression of the Sibme Platform: A Timeline
A decade ago, uploading a video to the internet was challenging, especially on mobile. Our initial mobile application was one of the first to successfully upload a one-hour video to the cloud from a mobile device in 2013. Since then, we’ve updated our web and mobile apps to make recording and sharing videos in Sibme faster than any other application. We soon found that, although many educators realized the value of watching videos, viewing whole lessons took too much time. So we developed ways to get to the most teachable moments in a video as quickly as possible.
Many teacher-student and student-student interactions occur within a lesson, and it’s easy to be distracted by superficial things when you watch a video of yourself (or others).
With the addition of Time-Stamped Comments, Sibme users could highlight important moments related to a teacher’s learning goals in a video. Viewers could stay focused on the most important moments in a lesson, making conversations more productive.
Rather than watching a 40-minute video, educators could instead focus on the 10 minutes that included comments. The feature also allowed time-saving asynchronous conversations between educators and coaches.
However, Time-Stamped Comments initially required someone to watch the whole video. And even with 2x and 3x playback speed and auto-pausing while typing comments, providing meaningful feedback in comments was laborious.
Synced Notes and Sibme Live
We believe that it’s important to have clear evidence for teachers and coaches to reflect on together, and two new features provided ways to make the process more accessible and streamlined.
The Synced-Notes feature allowed the observer to simultaneously record a video and add time -stamped comments while in the classroom.
Sibme Live lets someone virtually observe a classroom, commenting on the lesson as it’s happening live.
Docs/Three-Pane View
If we could find a way for educators to attach lesson materials and student work to comments in a video, it would make it easier for them to get a complete picture of what happened during a lesson. And if we allowed them to comment on the video and the other learning artifacts within a single screen, it would further support and facilitate richer conversations to improve every aspect of a lesson.
We added the Three-Pane View feature to enable educators to view all materials in one place and navigate back and forth between comments, moments in the video, and the artifacts of teaching and learning in one place.
Formulating an improvement plan can be overwhelming once teachers and administrators watch a video and view comments. We realized that educators were spending time mining additional data from videos, including the number of students working on a task and the types of questions being asked. With Sibme Artificial Intelligence, educators can now pinpoint the parts of their lessons that need additional attention and quickly and accurately evaluate a number of components of their classroom.
The Latest Evolution: Sibme AI
We didn’t develop an AI solution to chase trends—we took a strategic approach, leveraging the latest technology to empower educators. Sibme AI’s features give instant feedback on key metrics like student engagement, teacher talk time, the types of questions asked by teachers and students and more. You can even get a measurement of how much wait time teachers allot for student responses. And because the data is captured automatically, you don’t have to spend time manually tracking and analyzing it yourself.
Sibme AI uses natural language processing to analyze the actual words teachers and students are saying in a video. This means that you can get insights into the types of language used in the classroom, including the use of academic language, the level of rigor in the questions being asked and even the tone of voice used by the teacher.
Because data is presented in an easy-to-use dashboard, you can quickly recognize trends and patterns in your teaching practice or another teacher’s practice without having to rewind the video or take notes. If you already use the platform, you can easily combine AI data with the time- stamped comments, resource attachments, and other evidence of practice to get a complete picture. Data from Sibme AI is presented in an easy-to-use dashboard and in various reports. Insights from Sibme AI include:
Teacher talk vs. student talk: an analysis of when and who is speaking, when there is silence, and who speaks at any given time in the classroom. (There’s a privacy option that lets you blur out students’ faces.)
Question distribution by time: an analysis that shows when the teacher and students asked questions. By studying the height and distribution of these bars, you can gauge student engagement and conversational dynamics in the classroom.
Question distribution of close-ended vs. open-ended questions: an analysis of the number of open -ended vs. close-ended teacher questions. By comparing the frequency of each, you can understand the balance and nature of teacher and students’ questions during class.
Fleisch-Kincaid Grade Level: using a mathematical formula including average sentence length and the average number of syllables per word, the grade level communication a teacher is using is identified. This ensures students receive instruction at an appropriate level.
Words per Minute: an analysis of a teacher’s speech rate in a classroom video or audio file, helping assess delivery and student engagement.
Video or Audio information: a quick summary of the audio or video recording transcript in a paragraph. Without having to watch a whole video, you can get a quick synopsis.
AI Makes Human Instructional Coaches Even Better
Human beings are notoriously bad self-reflectors. It’s challenging to accurately remember something that happened in the past. You need to rely on some concrete evidence—like video—as you reflect. When you pair analytics with video, you can get an even sharper picture and a very high level of accuracy.
Teachers can use Sibme AI to better understand what they’re doing in the classroom and gain insights into how they can better engage with students.
Coaches can use videos and data in coaching conversations to ensure that administrators give their teachers appropriate feedback in observational conferences.
Administrators can use Sibme AI to reflect on good teaching and share best practices.
Even with automation in the mix, we still believe the experts should determine what to do next. AI augments human coaching. It doesn’t replace it.
When teachers can self-identify areas for improvement, they can better reach every student in the class, boost equity and greatly improve student learning outcomes.
What Sibme AI Users are Finding
Dr. Tonya Goree, Assistant Superintendent at Cypress Fairbanks ISD (TX), shares her district’s experience: “Sibme AI has been a catalyst for change in our district,” said Goree. It has significantly reduced the administrative burden on teachers and is providing some robust data to inform personalized professional development here. The platform also helps scale instructional coaching across multiple campuses, thereby democratizing access to personalized and actionable feedback.”
At Sibme, we’re all about backing the real heroes in our schools, our teachers. This is our way of saying, “We’ve got your back.”
About the author
David Wakefield is the Founder and CEO at Sibme.
