Flipped Classroom Pioneer Launches New Instructional Framework 

Jon Bergmann, one of the pioneers of flipped learning and heavy promoter of mastery learning has created a new teaching framework he’s calling The Mastery Flip, which, as you've probably guessed, adds AI to the mix. 

As a quick refresher: 

• Flipped learning is moving the acquisition of knowledge outside the classroom, through readings (analog or digital) and other digital mediums (videos, podcasts, etc.) and using class time for active practice.  
  

• Mastery learning acknowledges that not all students learn at the same pace and suggests that they should not be assessed at the same time. Allowing students multiple chances to demonstrate their mastery also promotes a growth mindset.  

Bergmann has three pillars for this new framework: 

Pillar 1: AI Engines (Turbocharging the Independent Space) 
Pillar 2: Analog Roots (The Anchor of Learning)  
Pillar 3: Human Checks (Validating the Cognitive Journey) 

Pillar 1 is the flipped classroom - now supported by AI. The major innovation here is that with proper development, each student could have access to a little professor-in-a-box that can help guide a student back to material that needs review. Here’s an example: 

Student: “I don’t understand <topic X>.” 

Prof-bot: “Topic X was discussed at timestamp 2:35 in the 2nd video we covered. Why don’t you tell me what part you do understand? Or we can review it again <link to video starting at 2:35>” 

AI is not yet reliable enough to trust this on its own, partially because of “hallucinations” but also because we’re not accustomed to questioning what is presented to us by an AI (which often feels definitive or authoritative). An AI chat bot trained to respond with guiding questions might be less likely to misinform. 

If the AI captured each time a student asked for additional help, alongside its response, it would be possible to improve the AI tool by offering corrections back to it. An instructor could review and offer corrections, improving instructions for the AI tool, which in turn leads to better responses to students’ questions. An AI chat bot programmed to watch for patterns of common questions can lead to gaps in the learning material, which can then be improved by the instructor. 

Pillar 2 is in-class time for review of students’ learning and course correction (pun fully intended). This pillar is built on the in-class time that allows instructors to see students practicing new skills and knowledge gained during the flipped portion of the class. Questions asked during pillar 1 can be summarized to offer additional guidance to instructors on what to watch out for during a class. Specific topics or questions that were asked, especially if there’s a pattern, can help instructors structure their in-class activities.  

Pillar 3 is the assessment portion, and it has different challenges. I believe part of assessment, particularly formative assessments, can be automated to some degree. The same material used for pillar 1 can be used to build better assessment tools – not to be a final judge on a student, but to help guide both the student, and the instructor, to where a student needs the most help. 

The same kind of programming that might detect patterns in questions during independent study can watch for patterns in assessment results that might not be noticed by an instructor or several TAs with different sets of students, giving specific feedback about where students seem to be struggling.  

This feedback can prompt an instructor to check why that might be. If assessments use different versions of the same question, this can even help you determine if the wording of one a question may be confusing or misleading for students. 

Much in the same way the chat bot could guide the student back to material in Pillar 1, it can do the same here for instructors, queueing up material for the instructor to review. 

Here’s one potential example:

This method might allow an instructor separate the more rote type of knowledge from the critical thinking portions of assessment. However, because of the current limitations of AI, I do not recommend its usage for summative assessments. 

Final Thoughts 

Jon Bergmann’s Mastery Flip model provides an opportunity to reimagine what education professionals can accomplish when they have access to richer data and more time to focus on what matters most. AI tools, when built and used purposefully, can shift repetitive brute force work to automated systems, returning time to instructors and freeing them to engage deeply with students – creating more effective learning environments.