In this newsletter, we’ve talked about lots of issues around AI in education.
Lots of them deal with instruction — how AI fits into classrooms, into teaching and learning.
But there’s a whole different side that can have a huge impact as well …
How AI can help us analyze data and make better decisions.
Of course, there’s still a LOT of heavy lifting for the human brains here. We don’t want to feed AI a bunch of spreadsheets and PDFs and ask: “What should we do?” … then mindlessly do it.
But data has been a barrier for lots of us — because analyzing it is time-consuming (and, if we’re being honest, we sometimes don’t feel like we know what we’re doing).
In this case, AI can act like a junior data analyst, giving us what we need so the humans can chart a course forward.
Today is the first in a two-part (maybe three-part … maybe more?) series on using AI to get insights from data. And I need your help!
Please answer the poll below — and give me your examples, tips, cautions, questions — about using AI to analyze data in education. I’ll use it in next week’s newsletter!
(Or if you prefer, just hit reply and send me an email. That ALWAYS works.)
In this week’s newsletter:
✅ Are your students doing test prep the right way?
📚 New AI resources this week
📢 Your voice: Student thinking in the AI age
🗳 Poll: Analyzing school data with AI
📈 Using AI to make sense of data in schools
✅ Are Your Students Preparing for Standardized Tests the Right Way?
This message is sponsored by Knowt
Your students are already using Knowt — and your teachers are starting to notice. Knowt turns everyday classroom materials into consistent, effective test prep: flashcards, study guides, and practice tests built from the content students are already learning.
📊 Clear visibility into progress: Teachers can track student mastery without extra training or onboarding. Adaptive practice and spaced repetition adjust to each student’s needs.
🏫 Admin-level insights: A centralized dashboard gives district and school leaders a clear view of adoption and engagement across classrooms.
📝 Exam Hub: Targeted prep for AP, SAT, and ACT, with practice questions and mock exams designed for test day.
🔒 Secure and compliant: FERPA and COPPA aligned to protect student data.
📚 New AI resources this week
1️⃣ 3 Ways Students Can Use AI to Improve Literacy Skills (via eSchool News) — A practical look at how AI can support literacy development—paired with a strong message that schools must explicitly teach appropriate use, not avoid the tools.
2️⃣ The Ed-Tech Backlash Is Here. What It Means for Schools (via Education Week) — Growing parent and teacher pushback on screen time is forcing leaders to better justify and refine how (and why) AI and tech are used in classrooms.
3️⃣ Student Panel: AI Is Here to Stay—But Don’t Replace Teachers (via Education Week) — Students themselves are calling for balance: they want AI tools, but not at the expense of human teaching and relationships.
📢 Your voice: Student thinking in the AI age
Last week’s poll: How do we keep students thinking in class in the AI age?
🟩🟩🟩🟩🟩🟩 Base teaching on pedagogy (14)
🟨🟨⬜️⬜️⬜️⬜️ Emphasize special human skills (5)
🟨🟨🟨🟨🟨⬜️ Use AI to support/strengthen critical thinking (13)
🟨🟨⬜️⬜️⬜️⬜️ More of the 4 C's in the classroom (6)
🟨⬜️⬜️⬜️⬜️⬜️ Other ... (4)
Critical thinking: “We have to teach them best practices. The problem is that the teachers are still being taught best practices. Creative teachers have to step up and lead the others into this new age of learning. We are all looking for the best ways to use it to strengthen critical thinking.” — T. Applegate
Critical thinking: aaa “I think we need to be more cognizant of designing lesson, activities, tasks that scaffold student thinking and writing and talking in a old school way and then transition to AI as a new tool for feedback, research, and alternative ideas. Cooperative Learning structures built in to lessons that purposefully provide parameters, limit resources, encourage human-self thought and reflections followed up with human-human exchange, and then human-AI collaboration. Highly effectively lessons moving forward will be designed to hit all those dimensions to better ensure actual learning.” — M. Tornetto
Other: “With and without AI, need to use more questioning and conversation strategies to get kids critically thinking about what they hear, see and read. I look at what our SLP does to get kids using visuals to connect to vocabulary and to discuss and analyze a photo or simple sentence. It’s crucial because in the day and age the students are lacking the deep dive skills that are even more necessary in this AI world.” — S. Plante
Other: “I think we need to build lesson plans/unit plans, instruction so that we focus less on the final product and more on the thinking and learning. It's messy, but why can't we look at the messy? That involves 1:1 conversations with students, having them make notes of their thinking, questions, changing of positions, providing feedback on the messy process of learning.” — C. Sussex
What would you like to read in AI for Admins?
What’s a topic you’d like to see covered here? Hit REPLY to this email and let me know.
Have you done anything you’d like to share with the AI for Admins community? Hit REPLY and let me know.
Would you like to write a guest post to support and equip AI for Admins readers? Hit REPLY and let me know.
🗳 Poll: Analyzing school data with AI
Instructions:
Please vote on this week’s poll. It just takes a click!
Optional: Explain your vote / provide context / add details in a comment afterward.
Optional: Include your name in your comment so I can credit you if I use your response. (I’ll try to pull names from email addresses. If you don’t want me to do that, please say so.)
In your opinion, what's the most powerful way that AI can analyze data in education?
📈 Using AI to make sense of data in schools
AI can turn data from overwhelming to insightful. (Image: Google Gemini)
Have you ever seen the huge binder of data?
I still envision the three-ring binder … maybe the 2” variety … being dropped on my desk with a resounding “thud” and someone asking: “Hey, can you crunch the numbers and let me know what you think?”
It sounds so innocent. To the people who make this request, it doesn’t sound like much.
But we know that, for teachers and admins and everyone else, “crunch the numbers” is time consuming.
Data has the potential to unlock insight into TONS of important things — changes to make, discoveries about individual students or an entire class of students, trends in the world that impact us.
When faced with a huge amount of data, it’s easy to think: “I don’t even know where to start!” … or “This is so enormous and overwhelming.”
Thanks to AI tools like Google Gemini, ChatGPT, Claude, and NotebookLM, there’s a lot we can do.
Today is the first in a two-part series (maybe more?) about using AI to analyze data and get insights into steps we can take.
FYI: I’d LOVE to pull in examples of how you (and subscribers like you) are using AI to analyze data and use it. Please answer the poll above — and describe what you’re doing in a comment!
Tips and safeguards when analyzing data with AI
If we aren’t careful, we can put the people we serve at risk if we aren’t careful with the data we’re using.
⚠ CAUTION: Here are a few steps to take before you start analyzing data with AI …
Strip out any PII (personally identifiable information). Before feeding any kind of data to an AI model, remove names, email addresses, student/staff IDs, social security numbers … anything like that.
Understand how the AI model handles data. When you use free versions of ChatGPT and other AI models, YOU are the product. They’re gathering the info you feed it to train the next AI model. However, certain tools — like NotebookLM — tell you that the data you give them won’t be used to train future models. Also, if you’re using a school Google or Microsoft account, your data likely won’t be used to train the model. (Example: In Google Gemini, the shield in the text box (see image below) shows that data won’t be used to train the model.)
Determine how sensitive the data is. Maybe this is just me, but even if the tool says your data won’t be used to train future models, there’s just certain sensitive data that I don’t want to leave my/our control. Example: discipline records, counseling records, health info, etc.
The shield in Google Gemini shows the data won’t be used to train the model.
Types of data to analyze using AI
So … what kind of data can you analyze using AI? Here are several ideas …
Classroom data — quiz results, test data (from state/standardized tests but also chapter tests), data from online practice platforms, assignment completion data from your student information system (SIS).
Surveys/forms — Have students (or teachers … or anyone) fill out a survey to gather data. Ask the questions that elicit the data that you want. Then, use AI to look for common themes and suggest next steps.
Student data — Again, no PII, but data analysis might give you a better view into your student population — as well as their households and communities. You might look at attendance patterns, behavior trends, trends in course performance, etc.
Community/state/national/international data — Download (or copy/paste) datasets from government sites and analyze. This can help you better understand the people around you and spot larger trends that could inform what you do.
School/district data — Anything specific to the school or district, like: school comparisons, longitudinal trends, how you’re allocating funds and other resources, etc. If you don’t know what the data is saying — or what to do with it — AI can help clarify. (But remember: the decisions are still in the hands of the humans.)
School/instructional initiatives — Any data you gather around these initiatives (like PBIS, RTI, etc.) can help you see how they’re progressing and adjustments to make. Of course, you have to know the strengths/weaknesses of the data and what they do/don’t show.
How to analyze data with AI
Once you have some data, what do you do with it? How do you get insights and suggestions? This is just a starting point …
Use the built-in AI. Some platforms automatically analyze data and provide insights. Just use it … look at it … turn it on and see what it says.
Use AI tools. We’ll dive deeper into this next week, but using a major AI model like ChatGPT, Google Gemini, or Claude will take you far. (Reminder from earlier: the free versions likely use your data to train future models, so be careful.) Other tools like NotebookLM can help you visualize data and create products with it to share with others.
Export data. With some platforms, there’s an option to export to a PDF file or CSV file. (CSV = comma-separated values … it’s like an Excel/Sheets spreadsheet but simplified and easy to read by AI.) Again, if you’re going to export data, be sure to strip out PII (personally identifiable information).
Think about the data you’re using. It helps if you understand what data you have and what you want to learn from it. Sometimes people just download a file and say “tell me what the numbers are saying” or “what should I do with this data?”. That works sometimes — if the AI has enough context. The best scenario, though, is that you have a clear problem to solve — or question to answer.
Copy/paste works, too. If you find a table of data on a website, use the highlight/copy/paste technique. When we paste that data into a document, sometimes it comes out a mess … but AI models are really good at sorting out and understanding data like that — even if the formatting is a mess — as long as it understands what the data is and some context around it.
Questions to ask with AI data analysis
When you have a fistful of data, it’s not always the data that unlocks action and next steps … it’s the questions that you ask that can be powerful. Here are some suggestions …
What patterns or trends stand out in this data?
What might be causing these trends?
What are 3 to 5 actionable next steps?
What additional data would help us understand this better?
What questions should we be asking that we aren’t?
AI can lead to better human leadership decisions
We aren’t outsourcing our leadership decisions to the robots.
We’re taking the time-consuming, often confusing work of analyzing data … and getting some help understanding it.
When you have better information — and a better understanding of that information — you can make better decisions.
In this case, more data isn’t necessarily better. It’s getting the data that matters — and understanding what it’s telling you so you can take appropriate action.
And when it comes to that action, AI can make suggestions on those actions. But it’s up to the human leaders to know whether those actions actually work for the other humans that they serve.
How have you used AI to analyze data?
FYI: We’re just getting started here! Next week, we’ll look into specific examples and specific tools that you can start using yourself.
In the meantime, I’d love to share some examples from you (and subscribers like you).
❓QUESTION: How have you used AI to turn data into action — and change? (In small ways or big ways.)
Two ways to share what you’re doing …
Fill out the poll (above) about using AI to analyze data. Share what you’re doing, your tips, your cautions, ANYTHING at all in a comment. (And add your name so I can attribute it to you!)
Hit reply. Prefer to just email me directly? You can always hit reply and reach me — for this and for anything else.
I hope you enjoy these resources — and I hope they support you in your work!
Please always feel free to share what’s working for you — or how we can improve this community.
Matt Miller
Host, AI for Admins
Educator, Author, Speaker, Podcaster
[email protected]
